Applied Materials Aktie

Well, good morning, everybody. I'm Mike Sullivan. I'm the Head of Investor Relations at Applied Materials. And once again, I have the honor of being the first to welcome you to SEMICON West. It's my pleasure. We really thank you all for being here. Applied has a very long tradition of hosting a technology breakfast at SEMICON. We've been doing it for years, and we've been looking forward to seeing you pretty much all year and spending time with you.

A big focus of this year's SEMICON, of course, is really understanding what is the state of the AI build-out and what does it mean for companies all across the entire technology ecosystem. And the figure of merit that everybody is focused on is tokens per second per watt. And when you look at today's presentations, what you'll see is that ultimately, AI -- every AI system and every AI chip depends on 2 ingredients, made of 2 key ingredients, and those are transistors and wiring. And what you'll see today is that over 60% of the transistors and wiring, 60% of those materials are put there by Applied Materials. And what that means is we have an enormous opportunity and responsibility as a company to make them even faster and much more energy efficient. We know the entire AI ecosystem is waiting for breakthroughs and inventing those and accelerating those. It's a challenge that we love. It's a challenge we accept. And you saw a video a second ago. Yesterday, many of us were in Sunnyvale, California for the unveiling of the Epic Center.

All of the world's top chip makers were there, Intel, TSMC, Micron, Kioxia was there, Hynix, Samsung, all there talking about how they're going to be working with us to develop these innovations and accelerate them and bring them to market together. So we're extremely excited about that. And so what I wanted to do just now quickly is give you today's agenda. Traditionally, what we do at SEMICON West is we focus on our core semiconductor products group, and that's Prabu, Raja and his team. And of course, they're here, and they'll be speaking. But this year, we're doing more. So we've invited Gary Dickinson to give you an update on the state of the industry and on our strategy. In addition, we've asked Keith Wells to join us this year for the first time, and he's going to give you insights into our fast-growing process control business. And we've asked Tim Dean to come and explain how we're accelerating the growth of our services business.

Usually, at SEMICON West because we're in quiet period, our CFO is kept hidden in Santa Clara. But this year, we've invited Bryce Hill to talk about our financial progress as a company and our plans and targets for the future. So we're going to be giving you a lot of new data today. Then all of our presenters, they're going to take the stage to answer your questions. And I've got 2 final quick housekeeping items for you. Number one is, as I mentioned, we are in our quiet period. So if you have near-term questions, we're going to answer those for you by next month on our earnings call. And finally, I've already had people saying, where is the presentations, where are the slides? I can't find them on your website. Please know that you don't need to write anything down. We have a tremendous amount of new information. A lot of the forecast that we had in the past have changed.

Many of them have increased. So there's a lot of new data. There are new financial targets. All of those are going to be available on our website as Brice concludes his presentation, that's going to be at about 9:45 a.m. So please relax, -- you don't need to write anything down. I promise you it will be there at 9:45. And that's about it for me right now. I'll come back and help you with the Q&A. But in the meantime, thank you again. It's so great to have you with us. And I'd now like to welcome to the stage, Gary Dickerson.

Well, thanks, Mike, and thanks for all of you joining us here today. This is an incredibly exciting time for Applied Materials. So many of you probably remember, 9 years ago, -- this thing is -- there we go. 9 years ago, we had our AI design forum and many of the top innovators in the data center were at that event. And over the last decade, Applied, we anticipated that AI would be the major driver of the industry, and we've been creating enabling capabilities to position us to grow quickly going forward. And that's what you're going to hear today. So AI is driving the global economy. It's driving the semiconductor industry.

The fastest-growing wafer fab equipment segments in the industry, driven by AI, our leading-edge foundry logic, DRAM and advanced packaging. And in all of those areas, Applied is #1 and positioned to outperform. And you'll see that today in the subsequent sessions. And one thing that I've been in the industry for a long time as an innovator, I'm engineer. I love innovating with all of our partners. The -- always there's a focus on technology innovation and performance innovation. And especially for AI, innovating in chips and packages for tokens per second per watt is a race for every one of our customers. They're racing for every new inflection to be first to market with design wins.

And so we have these incredibly deep relationships. I'll talk about that today with all of these leading companies. But the thing that I would say has changed more than I've seen any time in my career this year is the focus on capacity innovation, yield and output innovation. We're in an environment where demand is much bigger than supply. The chips have never been more valuable. So yield innovation, more good chips per wafer, higher-yielding packages is incredibly important, and that creates great opportunities for Applied Materials. We have been creating advantages that position Applied Materials with the most enabling technologies, and I will describe that in materials innovation, Mike talked about over 60% of the materials in the chip are coming from Applied Materials.

Integration innovation is incredibly important. Mike talked about EPYC, and we had all of the key R&D leaders through the technology stack on Sunday, yesterday morning, it was really amazing. We had hyperscalers. We had the top companies designing next-generation AI servers. We had all of the top chip manufacturing companies, design companies, materials companies. And one of the things that for me was very notable, every single one of -- and these are the top R&D leaders in the entire technology stack that were there, which is very rare, if ever, to have all of those people in one room. And what they said is it's not good enough to innovate on just components. you have to innovate from a system level. And we talk about high-velocity co-innovation coming together because the workloads are changing very, very quickly and the need for higher velocity innovation has never been greater than today.

So this integration innovation capability that we built at Applied Materials, certainly, it's co-optimization through that whole tech stack, co-optimizing with all of the technologies. Mike talked about how Applied is the leader in transistors, wiring, advanced packaging, again, co-optimizing across that whole flow is incredibly important. Applied has leading innovation -- integration innovation teams working 10 years out in the future, co-creating those future chip and packaging architectures from a system standpoint. Again, that's never been more important than today. And as I mentioned, good chips per wafer, yield in advanced packaging is incredibly important. And Applied has leadership in the highest resolution materials analysis. We're 2x bigger than our nearest competitor in e-beam, which Keith Wells will talk about later today, is incredibly enabling to accelerate innovation for transistors and wiring and high-speed memory and advanced packaging.

So these capabilities are tremendous. And then Tim Dean, who leads our services business, will talk about how we've created advanced service technologies with AI that will accelerate more good chips per wafer and more wafers out. And then we'll also talk about Bryce later will talk about value creation. Applied being positioned to grow very quickly in wafer fab equipment, over 20% increase in the coming years in revenue per tool in our services as we're accelerating R&D, enabling more good chips per wafer, enabling higher wafer output. Again, just for me, I'm meeting a lot of these top CEOs every month, top R&D leaders constantly. I probably will see all the top CEOs in the next 2 weeks. And that's because, again, they're all racing to drive performance leadership at these inflections. And I have never seen the pull stronger than today in yield innovation. The top priority for them is good chips per wafer and more wafers, especially again when the chips are so valuable. So you'll also hear about how we're growing our gross margins and operating margins through the day today. So I've been again in the industry for a long time. I've always thought what we do is incredibly cool.

I believe that we deliver miracles all the time in the semiconductor industry, but there was a period of time where there wasn't growth in wafer fab equipment. There was a period of time when you have the transition from 200-millimeter to 300-millimeter, then Moore's Law started slowing down and more of the growth was driven by power performance area cost and time. You see a 14% compound annual growth rate. We're not going to give a specific WFE forecast today. But what I would say is that I have tremendous confidence computing demand is going to grow at a tremendous rate. The compound annual growth rate is going to be very strong. So very confident in that. And you'll hear today also how Applied is using AI across everything that we do. The pacing factors are fab space, certainly supply chain components, energy, financing, some of those factors will impact the rate of growth. Value pricing can increase the rate of WFE growth.

But again, I have very high confidence as we're working across this whole ecosystem and technology stack, and I see how we're using AI inside Applied Materials, it is fundamental to our competitiveness, and I'll describe that also, and Brice will talk about this later today. You'll also hear from our Chief AI Officer, who started his PhD at 16, brilliant guy. We formed an AI leadership team 3 years ago, very clear ROI inside Applied, and we'll describe that. Okay. Leading edge foundry logic and DRAM compute memory. This shows the percent wafer starts for mobile, PCs and data centers. And what you can see in just a few years, there has been a tremendous inflection. A few years ago in leading-edge foundry logic, mobile was the major driver of wafer starts. Now data center has crossed over and data center is growing at a tremendous rate going forward into the future. In DRAM, you see the same thing, data center, high-performance computer really driving wafer fab equipment starts. So the great thing for Applied Materials is that in both of these segments, we are the leader. We are enabling the key inflections for performance and power and the rates for our customers for their design wins. And I think as all of you know, the chips in the data center are huge. They're at the reticle field limits.

So that drives tremendous growth in wafer fab equipment, and this is going to continue to accelerate in the coming years. New AI applications driving growth and diversification. And again, yesterday, we had key R&D leaders that are responsible for innovating with the next-generation AI servers. We had hyperscalers. We had really all of the companies that are leading the AI technology stack at this event yesterday morning. It was really amazing. And what they talked about was that workloads are changing very, very, very quickly. There is a tremendous need for improvements in token per second per watt or tokens per jewel, one of the R&D leaders talked about. And what we've seen from a standpoint of Applied Materials in calendar '26 is a significant growth in Agentic AI compute demand. And so we've increased our wafer fab equipment forecast going forward because of this increasing demand. And then on top of that, in the future, you have physical AI.

And on the right side of this chart, what you see is the mix from a silicon area perspective. So you're going from more GPU-intensive architectures to in Agentic, a more CPU-intensive architecture. Now for Applied Materials, we're very agnostic about this mix. What I will show is that more of these devices are moving to the leading edge faster, and that's where Applied Materials has higher market share, and it positions us well for significant growth going forward. Okay. This is an AI factory pod. How many people here know what a pod is? Okay, not too many people. So a pod is a modular building block for an AI factory. And so what you see here, this contains all of the components that go into a pod. And a Genty AI consumes about 2 to 4x more wafers than a generative AI pod. And then this next chart Basically, you'll see that the number of wafers is increasing, 2x and 4x, depending on the configuration.

This is really Ruben and Ruben Ultra. And then you look at the device mix per pod, 85% of the content of the Agentic AI factory is leading-edge foundry logic, DRAM and advanced packaging with -- from a standpoint of wafer starts. And so Applied Materials, what we've talked about earlier this year is about 80% of the incremental growth in WFE this year is in these 3 segments. And we see this growth continuing forward. So our position inside these 3 segments is very strong. We have, as I mentioned earlier, these deep co-innovation relationships with all of these leading customers. And we -- again, we have high clarity because we're co-creating these architectures with these leading companies. When I'm meeting next week, I'll be with CEOs and R&D leaders, when we're meeting, we're focused on the top 5 to 10 enabling technologies for next-generation technology nodes.

They're laser-focused on that. And Applied Materials has the majority of those enabling technologies. Again, we have integration innovators that are working in trusted, deep partnerships with all of these leading customers, enabling the higher velocity for these inflections. And the focus on improving the tokens per second per watt in the data center is driving more chips to the leading edge. That's what you see here, how long it takes for different generations of technology. And so what you see here, server CPU, GPU, networking, ASIC, all of those chips are moving faster to the leading edge where Applied Materials has a very strong position. So I talked about this earlier. The race for AI computing leadership is the most important race in the global economy in the semiconductor industry. And as I mentioned, there's this technology race with chips and multichip connectivity is incredibly important, how you connect all of these computing components together.

And Applied, as I mentioned, we have the deepest, the strongest relationship in co-creating innovation in these areas. And as I talked about earlier, never in my career have I seen a stronger pull from customers for more good chips per wafer and more wafers faster. And that's what's on the right side of this chart. And again, Applied Materials, -- over many years, we've been building leadership capabilities in materials innovation, uniquely integration innovation with our connected and co-optimized technologies, our integration innovators and our innovation platforms. That's exactly what EPYC is all about. I'll talk about that in a minute. And then yield innovation, I'll talk more about that. Most critical technologies for AI computing road map are shown here. Again, there is a race with all of these companies producing these technologies to be first to market with design wins. It is a very exciting race. We love co-innovating with all of these partners.

Applied is the leader across all of these different areas. So this shows leading edge foundry logic from 5-nanometer to 1.4 nanometer. You see mobile, high-performance computing and data center. And remember that data center is driving more and more wafer starts going forward -- and here, what you see is that wafer fab equipment per start is increasing node over node. You see Applied's addressable market growing to more than 50% through this time period, and you see our share of served market growing when Peru, which is 1 of the most brilliant and trusted innovators in the industry is up here next, he'll tell you how we are enabling these inflections, how Applied has built over many years, capabilities that are unique to enable these technology inflections.

And then if we look at advanced DRAM, same thing, what WFE dollars per wafer start served market growing to more than 60% and and applied gaining share in served market node over node. And this is maybe for the people that aren't deep into the technology, you have these different technology nodes. 6F squared is a certain technology for DRAM, compute memory, then you're going to 6 square CMOS bonded to the array, 4 square, 3D DRAM, all of these things, things -- the main thing, as you move to these future technologies, especially as you're going more vertical Applied is in a tremendous position -- those of you that have been around for a while know that we gained around 10 points of overall DRAM share over a number of years. And we are growing quickly in this segment and with these 3 companies that are all Epic partners, we have incredibly deep co-innovation relationships.

Again, over 60% of the served market in leading-edge foundry logic and in DRAM Applied has the most enabling materials innovation technologies. We have unique integration innovators. And for our customers, again, it is a race for who will be first to market with the best technology for design wins. So great opportunity for Applied going forward. This shows the WFE market segments, percent of total wafer fab equipment, the technologies inside those market segments from Applied Materials, and then our positions. So if you look at, again, enabling these major inflections enabling the race for better chips, better packages and deposition applied is about 2.5x larger than our nearest competitor.

In material modification, and Prabu will talk about this. Next, we are greater than 2x larger than our nearest competitor. In e-beam, which is 200x higher resolution than optical, you can't fix what you can't see. Keith will show how that business is growing over 50% this calendar year. And that technology is incredibly important to accelerate innovation and transistors and wiring and packaging, again, 2x approximately larger than our nearest competitor in e-beam. And we're a leader in conductor etch in foundry, logic and DRAM. So this enabling portfolio is why customers are working so closely with Applied Materials to co-innovate. And I can tell you, again, when I'm with these R&D leaders, when I'm with the CEOs, it is so much fun.

Again, I really believe we create miracles all the time. We make them nearly impossible, possible. And so having this technology portfolio that we have created over a number of years, positions Applied to be incredibly strategic for our customers and for fast growth going forward. So I've talked about the materials innovation. And the ability with these technologies also to connect everything together, that's important. I've talked about integration innovation with co-optimization of these technologies. And you'll see integrated platforms where you're taking multiple technologies and connecting them together under vacuum for better electrical performance, applied as unique in those capabilities. We're unique and the integration innovation teams that we have at Applied Materials. We built these teams over 9 years. They're world-class and engaging with all of our customers. And then the innovation platforms, I'll talk about Epic next, how that fits into this race and integration capability we have -- as I mentioned, we have the highest resolution, a material analysis capability with e-beam, and we have AIx.

AIx is actionable insight acceleration services that are accelerating good ships per wafer and wafer output. I'll talk about that also in just a minute here. So eBeam is a business for Applied Materials, high resolution, about 20x higher resolution than optical. You can see into advanced structures in ways you can't see any other way. Keith will share with you today many applications that are accelerating R&D, innovation with our customers and accelerating Applied's R&D velocity across our entire portfolio. So this is an incredibly important technology, strong pull from customers growing more than 50% this year and will continue to grow at a very high rate going forward.

In eBeam, again, Keith in his metrology inspection business, later today, he'll talk about 3 growth engines. Optical is growing over 50% this year. We have a lot of a pipeline of enabling technologies that we're bringing to market that business will grow fast next year and also going forward. eBeam, growing over 50% this year, also growing significantly. And this acceleration across the entire portfolio with this unique technology is incredibly important for Applied Materials. And again, all of this data, AI, you need actionable insight and actionable data -- this gives you massive actionable data that you can feed into AI-enabled servers, that's what our AIX platform is to accelerate yield, good chips per wafer. Wafer output and accelerate R&D.

The service business I have to say I love Tim, I love all of our team. Tim is fantastic. And he, again, is so deeply connected with our customers. All of them are in a race every chip is worth a tremendous amount of money. So more good chips per wafer, more wafers faster is really what we're driving in our service business. And we created this actionable insight acceleration server platform many years ago, we've been innovating with advanced sensors in our platforms and today, we have over 37,000 chambers connected inside customer fabs with AIx. The majority of those are remotely connected. And what we're seein and this is also where we're deploying these advanced service technologies, AI with predictive diagnostics -- we can match the chambers faster.

Tim will talk about that later. That's really critical for good chips per wafer. Again, there's many innovations that we're driving here. This business growing more than 20% this year in the mid-teens going forward, and again, just tremendous, tremendous value for our customers. EPIC, we have all of the top leading companies that are in the race for AI compute innovation as part of Epic. And what we do inside of EPIC is we are working with customers through the technology stack, partners through the technology stack many nodes out in the future, N+1, N+2, N+3, Ns 4 and co-creating these architectures.

I've already shown how Applied is the leader in material innovation, integration innovation, yield innovation. So our customers, they're bringing their top innovators to co-locate with Applied Materials innovators and for them, it's early access to these innovations. We have an incredible pipeline of innovations. Our teams are amazing in delivering innovations that are critical for these fast-growing segments that are enabling AI improvements. And so our customers want early access to those innovations and then integrate that into their devices.

So having those innovators co-located is a really big deal. Inside of EPIC also, what we're doing is we're running loops. We have over 100 test vehicles. We have customer reticles, and we can innovate in the transistor very fast. We can innovate in wiring very fast, in modules in high-speed memory or in packaging very fast. So again, for customers, what's really important is being first to market with leading next-generation technology. And then after that performance innovation in the chip and the package, they want to transfer that technology first time right into their R&D facilities and then ramp yield, good ships for wafer, packaging yield as fast as possible. It is a race for every 1 of these companies.

For Applied Materials, we're designed in earlier, with our products, systems and services. We also have incredible visibility to what matters, where to invest, where not to invest, because we're co-creating 10 years out into the future with this platform and with the relationships that we have with customers. And for the customers and for Applied Materials, we're driving R&D productivity and efficiency.

Okay, let's hear from customers Micron engineers and researchers will work side-by-side to the Applied Materials teams to co-innovate the manufacturing technologies needed to scale next-generation memory economically. By aligning our road maps and bringing our teams together, we can move faster from research to commercialization. This center is a milestone in on equipment R&D, bringing together global team makers, time companies and world-leading resorts to design the future of semiconductor AI is driving the need for breakthroughs at both the device and package level from faster transistors and better interconnects to advanced packaging. Meeting this challenge requires trusted partners working closely together.

And EPIC gives our teams the opportunity to work side by side on new materials and process technologies, the brightest mines from AMED and SKs, we tackled the toughest highest value inflections in the AI memory road map grows HBM, DRAM and that. combining this memory expertise with AMED equipment and manufacturing leadership is how we will jointly navigate the incredible complexity of AI remorse.

The next frontier of AI will be defined by how seamlessly we integrate compute, memory and connectivity into high-performance systems. By pairing applied materials, engineering and manufacturing capabilities, with Broadcom system and custom silicon design, we are bridging the gap between hardware requirements and packaging innovations. This will accelerate the path from concept to production for next-gen AI systems.

Okay. Again, all of our partners are incredibly excited about Epic and our ability to co-innovate faster together. So let me just spend 1 slide on materials innovation in adjacent markets. We recently updated -- I believe, in our last earnings call that this business is around these businesses are around $400 million per quarter in terms of run rate, and I wouldn't model anything near term beyond that. But let me describe these 2 businesses that are in this other category. Our display business is a little over 20% operating profit today.

We've announced recently this MAX OLED technology. And what's important here is this technology enables OLED to go to larger display sizes. IT, TV, automotive, and eliminates the fine metal mask, which has been the major technology constraint. We're working with leading display companies to implement this technology Again, the key focus here is the same thing that I talked about earlier is bringing those yields up to the tipping point for this technology to move into high-volume manufacturing -- so we're making progress. But again, I wouldn't model anything different today. Our SAM goes up significantly as OLED adoption increases. And 1 thing also about our display business. We are bringing our leadership and display deposition and eBeam into semiconductor panel packaging. And that will be worth a significant amount of money, not in the display business in our semiconductor business going forward.

And then the other area, just very briefly, very early on, again, I wouldn't model anything at this point. Applied has announced recently this since brand. We've created waveguide technology that has the leading optical performance at the lowest cost I can tell you there's a tremendous amount of innovation in this technology for AI glasses.

And around 90% of the flow to build these waveguides our applied custom innovations. Here, the strategy is to sell waveguides and eye pieces. But again, for both of these businesses right now, I would still model the roughly $40 million per quarter, and I think Bryce later will talk about combine these businesses in our future view are over $2 billion. They are multibillion dollar upside, but I wouldn't model this yet. And then, again, I -- as I said earlier, I've never had so much fun. I love working together with all of our partners. And next, Prabu will talk to you about our semiconductor business -- as I mentioned earlier, Peru is 1 of the most brilliant and trusted innovators in the entire semiconductor industry.

I talked to him probably 5 times a day every day, 7 days a week because again, it's just we love this innovation. So now let me turn this over to Prabu. Thank you.

I think it's always difficult to keep up the energies. I don't like to come after Gary. It's difficult to keep up the energy. I think if you really look at those -- if you are to look at those, I always think about, if I have to do 1 pad, what do we do, 1 page, my message is, we saw this AI inflection early. We invested in this influxion early. And we built our AI compute leadership over many, many years, over more than 10 years. Now it is the opportunities. It's a big tailwind for us. This is a simple message.

We built this readers. I spend lots of time with our customers and discuss on what I have very different -- now I'm going to tell you why I'm confident that we're going to outgrow the market, we are going to gain the share. That's what my presentation is about. Really look at those -- the foundation is a is all about power, performance and bandwidth. That's what it's all is about. The foundation for this AI compute is leading high-speed memory advanced packaging. When is a leading logic transistor and more wire, high-speed memory is DRAM, high-speed DRAM and a high bandwidth memory and packaging is all about chip-to-chip wiring.

Really look at all these 3 segments, the fastest-growing segments now today and like I said, we are #1 in each of these segments, and we are enabling this segment. That's why I'm confident about these 2 points or more than 2 points of share increase. So talk about advantages. What is your main advantage. Really look at those. Like I said, we saw the influx early, the investor in the early. That's why we are #1 in each of these segments. Leading ag logic, DRAM and advanced packaging. We built broad and enabling portfolios for these inflections, and we can connect them in unique ways to solve technology challenges, which no 1 else can do because of our broad portfolio, because we know how to connect them. We are building EPIC, new R&D collaboration model with our customers -- this is for faster innovation to customer commercialization and a higher innovation success rate. I will go through in details.

Our leadership is built over years, not today. If you really look at those, we invested in R&D centers. We invested in capabilities, and we built partnership over years. And if you really look at the R&D center, come back to R&D centers. We saw the packaging inflection much, much earlier. We invested in the packaging R&D center in Singapore in 2001. We saw the materials inflections, materials engineering, and started this material slab with the in-U.S. -- in Singapore 2018. Also, you can see the full flow lab in meta-centric now the EPIC, you take into a whole new level, all about high-velocity co-innovation with our customers.

That's what our customers know it. We also built unique capabilities with innovators who could connect all the way from materials to process, to equipments to device to systems. They understand better, they know how to connect it. And we build a partner see for complementary capabilities. basically is a good example. You don't have bonding and we built with -- for a die-to-wafer bonding we built a partners. And also we occurred companies. We know panel is coming. We are core companies, Tango, PVD panel next and also plating panel and also cigarette for x-ray and if relates provides us great foundation for our future growth AI is driving market enable WFE, the fastest-growing segments are fondilogic, DRAM and advanced packaging. We are #1 in each of these segments.

In the past, year, Applied Materials has outperformed the market in all these segments. In the next 5 years, the market outlook is even more favorable of incremental WFE growth will be in these segments -- Sorry. Yes. Like I said, it's all about power, performance and bandwidth. There are architectural inflections that will enable these devices, power performance and bandwidth in logic. You can see get on or out, better transistors, more transistors and more wiring, backside power, even more wiring, DRAM going through the inflection, high-bandwidth memory, even 6 scare, a lot of innovations left, 4 of current 3D DRAM for higher density and lower power packaging, system-level packaging, 1,000x more IOs and lower power. Applied has leadership in all these inflections.

Today, I will show you how we are growing our sham and share in these critical inflections. Really, they look at this. All the infants are going tread. One thing, if you look at common in last everything is going ready, which means more materials in generic. This gate all around transistor, example is a case study for materials engineering in Acton. Gary talks about magic. This is where it is happening right now. Really look at those. You think of it what a gate on cancer looks like in the picture -- in reality, how it looks -- this is how it looks like at an atomic level, or term by auto, how it is arranged. And I really look at those, where is the complexity between the 2 nanosate 10-nanometer space, 100 an storms, 10-nanometer. There are more than 5 distinct materials, 1 on top of each other. Each 1 of them want to 2-nanometer thickness.

Now when you look at those, the interfaces between these materials need to be engineered for best performance. interface dominate, interfaces are thicker than the materials itself. So you've got to play with the interfaces. That's where materials engineering coming in. Every angst matters, every no matter, even depositing it, modify, analyze every young star matter, and you need this atomic level precision at an industrial scale. That is commercialization. This is the power of our unique and connected portfolio.

Really, Gary talked about this slide. Applied has a broad portfolio across depositions, edge, modifications and metrology and the inspections. We are leading in many of these technologies, and we have a unique way of connecting these technologies for a better device performance. Like you saw on the last page, we understand this inflection better, and we know how to address them by connecting our technologies through co-optimization and integrations. Cooptimization where independent technologies are optimized for a better performance. You saw the last page, integration, where multiple technologies are integrated in 1 tool under vacuum.

You saw last page, different materials or more, the interfaces or 1 nanometer thickness, they cannot be exposed to yet. If it exports is gone, you don't see that. That's what need to be integrated. Here is that -- and also where we don't, however, we are extending these technologies to our partners where we have complementary technologies. These connected solutions industry's most critical challenges, example, co-optimization. This is 1 transistor contact. I can tell you, this is 1 of the most critical step in the device. Traditional staff, it takes 3 steps to do that. You will see that yet deposition and Epi deposition.

Traditional model, if you do, it limits the transistor performance. you don't get good performance, let me put this way. In the new corporate what we do, where edge CVD are improving their performance very significantly. The #1 bottleneck for our customers today is transistor performance. Device makers target 15% performance not over node. It's a full load, I'm talking about. This coat solutions, delivered 5% performance gain. This alone delivered 5% performance there. Value for our customers device performance. And yes, value for Applied.

Cumulatively, these products will deliver over $2 billion revenue for Applied over until 2030. Another example of wiring. This is why we consume 50% of the chip power today. As the wires become smaller, node over node, resistant goes up and further increase the power consumption. And our portfolio defines wiring road map. Our portfolio really defined worrying road map. We talked about the 3 solutions if you put multiple steps, they cannot be exposed to her. This integrated solution, this product is a good example. This integrated IMS solution reduced the resistant node-over-node.

I've been in PVD. This is for many, many years. Every year, we are the bottleneck for wiring. We find a way to remove the bottleneck and we make it happen. You can see node over node. These copper wiring products will deliver $3 billion revenue, 2026. This is just 1 of our integrated products. Our integrated solution portfolio revenue grew 20% to 30% of our system revenue today. Let me tell you how important Applied Material is for a compute. Over 60% of the materials on a chip today are created by Applied Materials upon. We have leadership products in transistors, wiring and advanced packaging, not just a new product going back to decades. In the last 12 months, we launched 13 new products for these 3 segments, transistor wiring advanced packaging. Apex takes it to a whole new level. If you really look at those, innovations from the universities on materials, all the way to the commercial chip, 8 to 10 to 15 years.

That's too long. It's a sequential process. It's very slow. It may not work anymore. In the Epic, what we do is a parallel model compressing the time 2, where you can see that materials products and development of customers all work together to solve a problem. So look at the customers and partners, universities, co-innovate with us under 1 roof, parallel development, value for our customer earlier access to materials, earlier relative to technologies.

This means for them faster innovation, faster R&D, faster ramp and faster commercialization, HVM. For Applied it will tell us that to invest R&D dollars efficiency. And more important, we are designed in early in our customer. Imagine these kind of complicated, complex structures. We are descending much earlier with our customers under advanced nodes. Now we talked about unique advantages. Let me talk about what are the big opportunities we have. AI is driving more demand for applied logic transistors unwiring technologies.

Why? Larger die, more transistors better transistors and more wiring. Our leadership in logic transistor and wiring. These technologies are moving to memory enough. Also, our leader in wiring technologies is extending to packaging. On chip wiring is moving to chip-to-chip wiring right now. And all of this will deliver strong SAM and strong Saad growth for Applied Materials. Here is an analysis of modern mobile chip versus AI training chip and agent chip. Three key observations. The logic silicon area transistor grows 70x from mobile to agent. The number of logic transistor grows 2x mobile to Ragentek. The number of wiring layers triples.

Remember, wiring is 1 of the Applied's largest business. where we have a very strong product. This is how AI compute is driving growth across our businesses. For decades, be innovated logic and wiring technology. I keep repeating logic and wiring technologies. AI is all about moving data faster and with a lower power. Because of that, our leadership technologies in transistor wiring are moving into memory right now. DRAM is more aggressive, most aggressive in adopting these technologies, transfer technologies, moving from planar to high metal gate, then moving to FinFET.

In DRAM, the number of wiring layers are more than doubling from conventional DRAM to HBM DRAM. Also remember, NAND is following transistor technology, DRAM, both in wiring. Wiring, in fact, it's more right now, you can see 5 to 10x. These inflection will increase applied salmon chat. Let me now show you how our SAM and share are growing in each of this node. I really look at the leading edge, the same -- the sham growth around, from 12 million note to 15 billion 1.4 nanometer and our share increases by 3 points over this transition.

In DRAM, we said in the past, that of Spire will be around $6 billion. Right now, from -- you can see that from 6 to 4 of Square, I think we are increasing -- increase -- and even for 3D DRAM, even more, revenues even more. And we expect to increase our sham share by 5 points. Wiring. I talk about wiring. Adrian and I can talk about wiring. We are #1 in on-chip wiring over many, many, many years. We are taking that leadership position from on chip wiring to the chip-to-chip wiring connections, high-bandwidth memory, 3D IC logic the same technologies we use to drive copper road map inside the chip are being adopted to connect the chips using micro bumps, you see bumps, using TSVs, using audios. Because of this, our advanced packaging is now expected to grow more than 5 billion within a few years. and we expect to outgrow the market in the next 5 years.

Let me give you 1 example of our leadership in the packaging in. Applied enabled 3/4 of the packaging process steps from TSB formation to micro bombs, both front side and backside of the wafers. We are #1 HBM packaging with more than 60% Sam share. HBM continues to grow, and we expect to grow over -- we expect over $1 billion this year alone. The next big packaging inflection is hybrid bonding. Really look at those people don't realize that there are many steps you can think of it as 3 groups. The process before bonding, the bonding itself and the process after bonding, we call postponing. Applied Materials portfolio addresses 80% of the steps across all these 3 areas.

In pointing, we partnered with basic where we have died to wafer bonding, we call it connect integrated tool. Our portfolio is just as strong in both pre and post bonding steps. Hybrid boarding is already being adopted in HVM, 3DC logic and also co-packagertics, the next big opportunities in HBM. Hybrid bonding will double our SAM as compared to microbumps and TSV. Another major inflection transition is silicon wafer interposers to panel. I repeated many times. Earlier, I told you we invested in packaging earlier 2012. That has given us leadership portion in wafer-level packaging. Today, we are doing the same -- exactly the same panel packaging.

Earlier, we saw the inflection early. We're investing in early. Applies already built a broad portfolio for panel inflexions including both organic and inorganic, such as recently we talked about but next for panel copper plating. And also, a few years back, we or panel PVD company. We saw this early, and we are investing early. We are very well positioned with our customer and confident to further grow our share. I will close today with a few words on our growth opportunity in ICAPs. An important part of ICAP, what we don't normally talk about, important part of data center, data center build-out is power and photonics. These are growing very rapidly, and we expect over $1 billion revenue this year.

Overall, I see that ICAP's market is growing this year, benefiting from 28-nanometer tranches. A lot of these nodes are transit 28-nanometer, where Applied is a really good push with the high-k metal gate transitions. In the future, we expect the physical AI to be a big growth driver for ICAPs. To summarize, AI is growing AF to much higher levels and applied SAM is around 55% of the WFE market. Next, our all of the AFF growth we expect through 2030 to be where Applied has #1 leadership positions, leading edge fondlogic. DRAM, advanced packaging. I say show you a priority is increasing both our SAM and our share in all these 3 segments, 3 areas.

As a result of this, Applied expect to gain over 2 percentage market percentage mark points of APL by 2030. See this growth includes our performance -- outperformance in metrology and inspection market, which Keith will now discuss. Thank you. Now over to Keith.

It's great to be here this morning. I'm Keith Wells. I'm in charge of our Image and Processing Control group, but you can kind of think of me as the M&I guy at applied. That's what I do. So we drive all the M&I business within Applied, and that's what I want to talk to you about today. So we have 3 thesises that are driving our growth in M&I. First is we see remarkable growth in optical inspection. We're -- I'm going to go into this in a little bit more detail in a minute. But this is a very, very big market. It's $9 billion. Today, we only serve about $2 billion of the $9 billion market with our products.

And what I set out to do when I came to Applied 6 years ago was to create a complete portfolio so that we could address the $9 billion in this market. And it takes a few years. This is not simple to make a lot of these technologies come to market. We won't really talk about these today, but I can assure you that in the next few years, you're going to see new products introduced in the optical market to go after this business and capture more of this TAM.

Second, thesis around our growth as we are the leaders in eBeam. And I'll give you many examples through the presentation today. Gary talked about it, Prabu talked about it. We'll show you in real concrete terms why we are the leader in eBeam with our technology. And then finally, you've heard it throughout the presentation today. We're going to talk a little bit more about how synergistic we are to help drive the semi systems business, and how that will result not only in growth for our business, but growth Probus business. So let's go into a little bit more detail.

So first of all, Gary already told you this, but in optical inspection, we grow 50% in revenue from 25% to 26%. We see very strong CAGR going forward in this segment. customers want alternatives, we're providing alternatives, and we're benefiting from that. We also see positions, as I said, in the segments of these markets that we don't participate in today. There are 7 billion of additional TAM out there that we can address, and we will be introducing new products over the next few years to address those. It won't be easy. But in some cases, this is owned monopolistically by other companies who fundamentally or customers are pulling very hard to -- they want solutions. They want alternatives.

So I think that's a very strong thesis to our growth plan. Second is eBeam. So again, 50% growth between 25% and 26%, mainly driven by the need to drive yields, very important. And we see strong share gains over the next 5 years in this segment. And what we did a few years ago is that we brought out cold field emission technology, which is very high brightness electron sources, coupled with high-energy electrons -- and by doing that, we were able to gain 2x in revenue over our nearest competitor. That technology, those architectures created differentiation and have really driven this business. And we intend to keep our leadership in electron optics by reducing many new tools over the next few years and introducing them.

And finally, the big theme here is how do we co-innovate. How do we work together with the semi systems with our AGS group to give them thousands of data points, so they can do experiments. They can increase the R&D pace by 3x. We now can do that in EPIC that will come online, we'll co-innovate. I'm very excited about this. We are the only company that has both advanced process, advanced metrology and the ability to test it in labs. There are no 1 else out there can take those 3 things and put them together.

I'm very excited about it. And what you're seeing is the financial results from this. We expect $3 billion in systems and service revenue in 2016 over that and then very high growth going forward. So that's our 3 thesis is. Now let's talk a little bit about the uniqueness. Gary presented this slide. And what I really want to do after I just recap the slide is to go into actual examples that prove our points around this.

So huge expansive portfolio, thousands of data points generated by our cutting-edge electron metrology and inspection capabilities going into our analytical platforms and the proof points you see over on the side. One is that we'll do $2 billion alone in e-beam metrology. So the market is voting for this type of capability. In addition, we see the acceleration of both semi systems and AGS innovation. And then as I said, again, finally, when we get to EPIC, I think it's all going to come together, and we're going to be able to move very fast, and innovate faster than any other company out there. But let's turn to some examples.

So Prabu in his presentation, talked about how important it is to have high-performing transistors. And so here is an example of how we enable high-performing transistors. Epi is key to the performance of the transistor. And when you do epi deposition, sometimes you get voiding, you saw it on prebudget. You don't want voiding. Now how do you determine whether there's avoiding there or not?

Well, what traditionally has been done is at. So they said, explain Tim, you're Gee, you need to tell people what Tims are. So what attemis, as you go into a wafer, you scrap the wafer, you cut a very, very small slice out of the wafer down deep into the architecture, you pull that out, you put it into a transmission electron microscope and you image. And you hope you cut in the right place, you probably get 5 to 8 transistors to look at. It doesn't give you much data. People love it because people love pictures and they feel very good about the picture they see there, and they can see the void in that image, and they go, okay, I need to work on my epi deposition recipes, very slow, very inefficient.

Now imagine you have an electron metrology tool that could go to thousands of points across the wafer and detect voice. You don't have to guess where to cut. You can create a full wafer map and that's what we're showing there. a heat map that shows you, look, there are voids in the red areas. They are not voids in the blue areas. So what do we do? We take that information, we start tuning our processes in order to eliminate the avoiding. We do that much, much faster than relying on Tims. And this is what's been driving much of our growth in this area because people clearly see -- they see -- and you can -- if you can see it, you can fix it. And not only is this used in R&D to accelerate R&D, it's also used after the fact in HBM to monitor the process.

So it's not just a business about R&D. It's also a business about putting this in line and monitoring it and then getting adoption with wafer starts, more good chips out. Here's another example that Babu gave and how do we help them. This is about wiring. So a very typical thing in wiring is obviously via fill and polishing, right? This happens many, many times in any process. You saw those high stacks that Prabu were showing you in his presentation. So what do you want to do? First of all, you want to deposit your metal and you want that deposition to be uniform and non-pattern dependent. And how do you figure that out?

Well, again, what we're able to do is after a PVD deposition, were able to go in and see, are you overfilling, are you underfilling -- or do you have pattern loading effects. You can now, again, with a full wafer map, thousands of data points go in and accurately tune your PVD deposition better than you ever could in the past. No tips. Then the next thing you need to do once you get the PBD done correctly is you need to polish. So you go in with your CMP and you polish it. And the problem is you can over polish or you can under polish. So you can have protrusions or you can have DISH. You don't want either. You want to get the polish right.

Again, the provision goes in there takes thousands of data points across the wafer and gives you feedback about how good your CMP process is. We put these together, we greatly accelerate our ability to deliver yield to the customer and ramp the process. And once again, this is an HVM application that people use to monitor the processes. So that's how we're helping transistors. That's how we help in wiring, how do we help in DRAM. So DRAM, as Gary alluded to, is adopting CMOS. But they're adopting it on an older node, a FinFET type device in most cases, instead of a gate-all-around type device. And so when DRAM customers came to us and said, "How can you help us make sure that we're accelerating our logic and transistor learning?" And we said, well, of course, we've done this in the past.

And this is an Epi example again. at slightly different Epi structures. But what you're seeing here is you don't want those epi structures to merge. You want a deposit such that they're clearly separated, so you get optimal transistor performance. And we said, well, several years ago, we put together the ability to image an open epi formation, which is what you want or a closed epi formation, which is what you don't want in order to get a good transistor. And we said we can deliver that to you in the matter of weeks. It's easy. We know how to do this. And so both in our labs and at major customers, we now -- they now use this to help tune their processes and look at reducing variability across the wafer. We often talk about yield good, bad die. That's not exactly what yields all about. Yield is also about minimizing variability.

So if you have transistors, some part of the wafer that runs slightly faster or slightly slower and another part of the wafer, that's bad. You do not want that. You want all transistors to run the same and the way you do that is to do massive measurements in order to reduce variability across the wafer. Again, another big driver of our business. Now this is a new market. This is exciting. This is where I feel like I can geek out a little bit. Imagine when you saw Pros cross-section of the atomic level gate all around. Imagine if you can go in and measure that non-destructively on the device. And that's what we created here.

So we took our CFE technology which is very, very high resolution, and we pointed it down at individual transistors and that will generate X-rays. Now if you put the right detectors around that, you collect the X-rays and you can actually identify very, very sub-instrumenfilm thickness and also composition. And you can do this on device and in line and so not only is it something that we developed in the labs over 2 years with our IMS group and Pabst area, but it's also something that we're now doing out of the customer. And why does the customer want to do this?

Because traditionally, the only way you understood whether or not you're making the transistor correctly is to wait to electrical test, which is weeks later, okay? You don't want to wait weeks. You want to make sure that transistor is performing correctly, the variability is low.

So while what we're able to do is to give the customer the ability to look at their transistor performance in line weeks before they actually have a functioning device. And so we think that this new market, which we're very excited about, is we're several hundred million dollars and it's not in our P&L today. It's growing, and I'm very excited about this. Now we all talked a little bit about advanced packaging in our presentations today. The thing about Avac packaging is it's all about yield. You're taking now and good chips things that yield it already, and you're going to put them into a very complex package. And the last thing you want to do is to destroy that known good chip and so the intensity of metrology inspection has gone through the roof in this market because people realize the value of known good package is out. And so a few years ago, our customers came to us and said, as these are getting more and more complex, what can you do in the eBeam area?

You've got even products in the front end line. Can you modify them and adapt them for the back of the line. And we did that and we're seeing revenue from that already across multiple products. But the other thing that came to us and said is, can you give us a cost-effective optical inspection solution. Some of these the optical solutions are tens of millions of dollars. We want an alternative. We want something that's far more cost effective. And so when we understood that, we started a completely new platform to address this market. This market went from a few hundred million a couple of years ago to now well over $1 billion because customers want to inspect every die for small particles before they place it, because a small particle in the wrong place, Onvia is going to destroy that known good die and destroy that package potentially.

So the value of optical inspection in advanced packaging is exploded. We're also not going to talk about it today, but we also have X-ray solutions in this area, too, that allow us to see very subtle effects in the bonding in the advanced packaging. So finally, just to recap. 5% growth in our core businesses from 25 to 26 with extremely strong CAGR going forward, okay? So very excited about that. Two, as I said, in new products, I came to Applied Materials. I worked at KLA for 27 years. I love metrology inspection. I overlap there with Gary. Barry and I worked together there. We're very passionate about this. What we wanted to do was to build out a portfolio. We want to have every product that can address the large markets in optical inspection.

Every product that addresses the large market and eBeam technology and then we've got some other things that we're pushing that I talked a little bit about today with advanced packaging, which we see extremely strong growth in and then also the material metrologies and other X-ray applications very excited about what we can do there. And then finally, what we're also doing is helping our partners within Applied to innovate faster, and to bring solutions faster and help with their growth. And it's extremely synergistic, which I hope you saw from the examples today. So thank you very much for having me.

Thank you. And now it's time for Tim to come on up -- all right. Good morning, everyone. Thanks for joining us this morning. It's a pleasure to be here. In Gary's presentation, he talked about AI computing leadership and how there are 2 paths to growth. The technology race, which Prabu and Keith just talked to you about and all the innovations that Applied is bringing to the market, and then the capacity raise, which I'd like to talk to you about now and how the services team enabled by our industry-leading AIx connected platform is helping customers optimize both yield and output with their existing installed base. This is the fastest way to profitability for our customers.

And then how as they bring in new capacity, we're able to ramp that capacity faster. So we're driving yield innovation, ramp innovation and output innovation, all within our AIx connected platform. So before I go into the details, I want to spend a little time on the graph here that Gary introduced earlier in the presentation. So many of you are familiar with the S curve for a factory, yield and output on the Y-axis time on the X-axis. And this outlines essentially the life cycle of a product as it moves through a factory. There's the R&D phase where we spend a lot of time with our customers optimizing their process as well as the hardware. There's the critical ramp phase where our customers are trying to, as quickly as possible, optimize yield and output and then there's the high-volume manufacturing phase where we're trying to reduce variability in the factory so that the fab is operating at maximum output and yield.

So through our AIx platform, we have created capabilities that help our customers in all 3 phases of this lifetime. And as EPIC comes online, you'll see that we're actually able to advance that even faster. So as fab utilization has improved this year, AIX adoption has increased this year. Our revenue has grown more than 20% in calendar year '26, and we've increased our future model to a mid-teens compound annual growth rate going forward. So EPIC is really going to change everything. It's not just about co-innovation of our new platforms with our customers in the collaboration areas of the facility. It's also about enhancing the co-innovation of our service products with our customers.

And how is this different? Because we have customers co-innovating with us in this facility on dedicated systems, we're able to connect our AIx AI-enabled servers to those platforms much earlier than we have been in the past. And we can monitor the tools as they're developing with the customers' recipes to get a good baseline fingerprint of the tools. Our service engineers from AGS will be operating hand-in-hand with customers in that collaboration space, which will allow us to get a good baseline on how these tools operate in production. And then when the tools are transferred from our facility in EPIC to either the customers' R&D or HVM facilities, we'll be able to ramp those processes faster, leveraging many of our capabilities, including crossmatch. So let me talk to you a little bit about how service works in the AI-enabled era. So as all things service starts with the tools. So we have an enduro here. It's got a lot of specialized sensors on the tools and all of our tools generate tens of thousands of sensor parameters, and we feed these into the AIx platform.

So what is that platform. It's a platform that focuses on actionable data tens of thousands of sensor means there is a lot of data. We don't need all that data to solve every problem, but we need to figure out which data streams are most critically important to solve the problem that we're working on. We run that through our Insights Engine, which is our analytical tools, our AI-enabled capabilities that are developed by our subject matter experts and we build that on 59 years of equipment design technology, which is our domain knowledge to create a suite of capabilities, which we use with our customers in their facilities.

As Gary mentioned, we have 37,000 chambers connected today. About 80% of those are very securely remotely connected capabilities in partnership with our customers we have more than 18 quarters of data that shows consistently that we get 30% faster service resolutions and even less escalations when we compare the same tool set connected to our AIx servers versus a tool in the same fab that's not connected to AIx servers. So what does this mean for customers? It means that as they look at all the variabilities that I mentioned earlier in the factory, whether they are tool matching variabilities, process recipe variabilities, the maintenance, health of the tools, minor part variabilities that might change the output of the factory, when we run them through our AIx innovation engines, we can tighten the distribution of variability across the fab. This drives more output from the facility as well as higher yields on the facility.

And I've highlighted a few of the capabilities that we've developed in the AX portfolio today. This is not an exhaustive list. This is just some of the key ones that we're working with customers on today. So let me talk about 1 of those in a little bit more detail, and that's cross match. So Crossmatch is a critical capability, and it works like this. When a fab and applied materials work together to optimize an R&D process, we've done it visually here by moving that gold dot into the center of the target on the right-hand side. So this is the best possible outcome that we can get on a given tool and process.

So we define that as a golden tool and a golden wafer. As a customer ramps through that very steep ramp period in the life of a factory, there are $1 billion of opportunities for our customers if we can ramp that fab faster, particularly in the current environment. So we take the cross match capability, and we move all the tools as we ramp them to overlap very nicely on top of the R&D developed process. And that gives us 1 golden tool in R&D moving to multiple golden tools in high-volume manufacturing. We've been able to execute software with our partners in the business units and the product side and we've been able to do this now 50% faster than we've been able to do in the past with this automation, and this drives more good wafers out and more wafers out of the facility for our customers.

So it's good for yield. It's good for output for our customers. It's great for us because it makes us much more efficient than we have been in the past in ramping these facilities and allowing our customers to drive to a maximum output. Let me talk about a couple of more examples that we have in the portfolio. Now before you look deeply into this slide, I want to explain what this looked like before we had the AIx portfolio. So this is an example where we have to replace an electrostatic chuck or an ESC over time in an etch process. For those of you that know this process, it's a very challenging environment in the chamber. There's challenging chemicals, which will eventually start to attack the edge of the eCheck, which is what holds the wafer down in the chamber, and you start to get a failure, maybe delamination at the edge of the electrostatic chuck.

Historically, we didn't know when this was happening until well after Keith tools would identify the problem through a review process and then they would go back and stop the tool and evaluate whether we need to do a change. All that productivity that was lost and good wafer outs were lost to that cycle time waiting for the metrology to identify the issue. And our first generation of our AIx solution, which is in the left panel, we were able to create a health index by which whenever we saw that challenges at the edge of the electrostatic chuck, we can immediately stop the process, and we wouldn't put that WIP at risk in the factory for our customers. But we knew that wasn't good enough.

We wanted to create a second-generation solution, which we see in the middle panel. We took our AI-guided models with our subject matter experts, and we created an algorithm by which we can clearly see how the life of the truck is progressing in the chamber. The little triangle shows when the failure happens. And again, we can now use a time-based algorithm to understand when that will happen in the future. The right panel is a real-life example with a customer on how we executed this change to our AIx platform with this Gen 2 solution.

So the triangles represent when the electrostatic chucks actually failed, and the horizontal blue bars show what our prediction model predicted they would fail. And you can see a very high correlation -- in fact, we now have the ability to predict 30 days in advance when a Chuck is going to fail with more than 90% accuracy. So why is this important? For a customer as they're looking at their WIP or their wafers in process and what material is coming to a tool, they can now take an unplanned event, which would be the failure in the past and turned it into a planned event, and they can time that when they have the best lull in their incoming WIP, so it has least impact on the factory.

For Applied, it's great because we have a very complex assembly here that we can bring in from our logistics system and we can have a trained engineer there, and we could do a controlled replacement of the chuck as opposed to an unplanned replacement of Chuck. It's good for the customer, it's good for Applied. Let me give you another example. This 1 got to advance here. This 1 is an ICAP example. So this is high-density plasma, HDP, CBD. We've had these chambers in the field for decades. Whenever we do maintenance, there are a number of nozzles around the inside of the chamber that have to be replaced over time.

The historical way we did this was we had open up the chamber, our engineer would replace the nozzles, they would close the chamber, pump it down, run a wafer, send that off to metrology, get the results, and then we have the graphic that you see on the left. An experienced engineer would then open up that chamber. And based on that map, they would replace the nozzles in a different configuration to try to optimize the nonuniformity on the wafer. They would close the chamber, pump it down, send it to metrology. This would happen up to 6x with an experienced engineer to optimize the nonuniformity on the wafer. With AIX Pro, we're able to do that now in a single review of the metrology map. So AIX tells us where we should position the nozzles.

We go back and we verify that and we take iterations down to 2 iterations, saves about 65% of the time for maintenance and drives more good wafers out for our customers. So it's a great outcome. Let me share a third and final example here, and this is yield innovation. Gary talked a lot about yield innovation and how important it is for our customers today. So in partnership with the equipment team, we've developed a unique sensor that sits on the front end of the etch system in this particular case. That sensor is connected up through the customer's manufacturing execution system and also connected to our AIX servers. And the purpose of this sensor is to allow us to fine-tune the placement in each of the etch chambers. And we connected to the MES so the customer can periodically direct the tool to do this recentering again.

And then, of course, it's connected to our AIx server so we can optimize that position with our algorithms. This is extremely important. You can see in the center panel that the edge effects that we see at the edge of a wafer really drive yield fallout. We get these tilted etches, which don't land on their contacts, they result in open and lower yield. In most wafers today, 6% to 10% of the die are on the outside edge of the wafer and yield is lower there. So if we can move those etches to completely vertical, we're able to make land more contacts, get better yield for the customer.

When we showed this data to the customer, they immediately fan this out across their entire installed base of these tools, and we have a second-generation solution in process right now. So I've talked to you a little bit about AIX this morning. I'd like you to hear from 1 of our customers, and that's Manish Mattia, who is the President and Chief Operating Officer of Micron Technologies. Please run the video. Micron is a leader in applying AI across semiconductor manufacturing, leveraging more than 1 billion monitoring points across our global fabs.

Our partnership with Applied has shown that meaningful results come from combining the best capabilities across both organizations. Applied and AIX leverage unique analytics service engineering expertise and predictive capabilities to solve problems that would have otherwise gone undetected. The collaboration has strengthened yield performance, improved productivity and created a powerful model for data-driven manufacturing excellence.

Micron's global adoption of AIX provides significant benefits for our technology development and manufacturing organization. and is a critical component of our valued partnership with Applied Materials. So Manish and Micron have been a great partner as we find AI out into their factories and appreciate him contributing to the presentation today. So I'd like to wrap up with a single oil here. And this is a summary of the business overall for Applied Global Services. You can see we had a long period where there was no growth in the business. And as we shifted to outcome-based services in the mid-teens, we were able to drive a low double-digit growth compound annual growth rate. But as we've moved into the AIX area, we've actually increased the model, and we think long term, mid-teens is the plan going forward.

You can also see that we've increased our subscription rates from the low 40% to nearly 70% in the plan going forward in 2030. And the model is really quite straightforward. As the installed base grows, and we're able to put more tools into our service network with AIx enhanced services, we're able to drive more benefits for our customers. And these benefits are in the form of our AIx services, as I communicated today, driving yield and output innovation. So we've got 37,000 wafers or 37,000 chambers connected today. We expect that to grow to over 90,000 chambers by calendar year 2030. And as we drive incremental value for our customers, better cost of ownership, we are rewarded with long-term contracts. They are currently greater than 2.5 years and at a more than 90% renewal rate.

In fact, our AIX services are statistically higher renewal rates than our service overall. And this is going to allow us to continue to drive profitability and a 20% revenue per tool growth between 2025 and 2030. So thanks so much for your attention today. I really appreciate it. And next up is Brice. Thank you.

Good morning. Nice to see everyone. Thanks for attending our breakfast, and thank you for your interest in Applied Materials. Today, we talked about how we're growing and enabling the most enabling portfolio in the industry. I'm going to talk about -- put that in perspective from an R&D perspective and then also talk about how we've delivered 10 years of excellent financial performance while executing that strategy.

We'll talk about how AI is accelerating our outlook and growing demand across the industry, we are investing to double our lab and manufacturing capacity for the company. We'll talk about our outlook, the targets we have for 2030, growing profitability and shareholder returns. And we'll also talk about our capital allocation priorities and how we will invest going forward, continue to reinvest in the business and continue to grow our shareholder returns. So on this chart on the right, we show how R&D grew over the 10 years from over $1 billion to over $4 billion.

Some people asked me when they're asking about Applied Materials, Oh, if I'm interested in the company, aren't I just buying the whole industry because you have a large portfolio. And I wanted to share that is not how we think about adding to our portfolio that we're adding scale or adding size. The way we think about it is what you heard today, being the most enabling company for our customers and for innovation.

And I'll just summarize a few of the examples here that we went through a couple of different color codes -- on the blue side, Prabu and Gary both talked about how the company's strategy shifted early on in this time period. And on this slide, we note the PPACt strategy in shifting from in the prior period, making the tools perform better, focusing on the operations of each tool, Shifting to actual chip performance.

So the PPAC TV strategy and platform, performance power area cost of the chip, the actual device not just an individual step that we were doing. Prabu talked about transistors. He talked about contact. He talked about wiring all the wiring layers that are in the chips as we work with customers to optimize those as bundles of solution from applied materials that can be reused by the customer, we're really able to help deliver a smaller form factor, lower power, higher performance, which is where you get excellent performance at the chip level from the customers.

At the same time, we also hired and built expert integration teams that understand how the chips come together. We built our Integrated Materials systems group where we put multiple tools together in a cluster so that you can build the component under vacuum. Prabu talked about how as we're assembling those materials, they never touch air and those produce some of the highest quality solutions for our customers. We also started the Metalla and the metal lab is where we can do short loops or longer loops. So if there's 22 steps in a process where you're building a part of a chip, we can do all 22, control that for the customer and build the highest-performing solution.

So that was really the first major change thinking about the inflection from Moore's Law sort of the shrink environment to no, you have to focus on materials. You have to focus on integration. The second 1 we highlight on this slide in the yellow is the company also understood that the way to advance system performance think about data center AI server systems is you have to put a lot of chips on a platform, on a package, on a substrate. Think about GPUs, CPUs, accelerators, network, stacks of HBM memory, all of these have to go on a package. And so we talked about the Internet capabilities.

We talked about panel processing. So the yellow investments that you see on this page where we have a partnership with BSE on hybrid bonding. We recently made acquisitions that help us with X-ray technology and thin interconnects on the packaging, and we had our Tango acquisition, which helped us with our panel platform, continued to focus on what do we need to have in the portfolio to be the most enabling for our customers to solve that next generation of performance challenge.

The last 1 we highlight here in gray is the AI investments that we made, the AIx platform is taking the massive amount of data from our tools and producing solutions for our customers that help them drive the tools to the highest yield and highest performance in their factories. We call that actionable insights, and it's -- we continue to add sensors and capabilities to our equipment that give that amount of data that we can use to drive performance for the customers. Keith also talked about how we significantly increased the product road map in our metrology and inspection business, and that's so we can the problems that need to be fixed in our lab, and we can also serve the customers with the equipment they need to see the same things in their HVM.

Finally, some of you saw that our EPIC center that we just unveiled yesterday. EPIC is where we bring all of these solutions together under 1 roof to work with our customers on the long lead challenges that the industry had, and hopefully, you had a chance to see that. So with all of this R&D, we feel extremely well positioned as we look forward to new technologies and process technology and enabling our customers.

So 10 years of excellent financial performance. The graph on the left, the bar graph shows the revenue growing over the 10-year period. at a 12.2% revenue CAGR to $34.3 billion. Gross margin on the top line grew 680 basis points to 50%. And over the time period that's non-GAAP gross margin. Non-GAAP operating margin, up over 1,100 basis points over the same time period, earnings per share CAGR, 22%, nearly double the revenue CAGR over the time period, benefiting, of course, from buybacks, which we'll cover later, average free cash flow. So the business is a high profit cash flow business. And the last 2 points are really important, I think, for investors and investors that may not know the story as well as others. We can reinvest in this business.

We're reinvesting more than 50% of our cash generated back into the business, R&D, CapEx, working capital, small acquisitions to grow our footprint grow our solutions, grow our service capabilities and continue growing the business. So greater than 50% reinvested and -- there's a greater than 30% return on invested capital that we've averaged over these 10 years, and we expect that actually to be increasing as we move forward while we continue to invest in the business. So executing the strategy of building the most enabling portfolio that we described over 10 years, delivering excellent financial results. I'll 0 in on gross margin, of course, there's a lot of interest on the path of gross margin it's indicative of the strength of your product line, which we say is becoming more and more enabling.

So 680 basis points over the 10 years the drivers, if I break out the drivers for the change and increase in gross margin, the largest 1 and what I want investors to think about most is the portfolio itself, more and more enabling. Every time we build a solution that is critical and helps the customer innovate. We spent a lot of effort on that, and it's providing value to the customer. And if we do a good job on pricing, we've worked hard on our pricing process for the last 4 years. If we do a good job on pricing for the value that, that tool is providing, then you see the value of the portfolio increasing, and that's the #1 driver of the increase across this time period.

Number two, our semiconductor systems business, that's our largest business. It's our reportable operating segment. We report our gross margins for that business. It's over 55% in the recent quarter. So as that business is growing faster as the business has accelerated, that gives us an uplift from gross margin. And finally, we have operational efficiencies -- we've diversified our manufacturing base. We have a large operation in Asia, in Singapore. We've diversified our supply chain that helps us grow and scale our profitability, and that's helped from a gross margin perspective also.

Now as we look forward and we get to the end of the presentation, we expect these same drivers to be in place. We expect growth in our semi systems segment. We expect to continue to innovate at the portfolio level and produce enabling solutions, and we expect to continue to expand our manufacturing operation going forward. So we think we're well positioned to grow the value from a gross margin perspective. Switching and thinking about the forecast for a second as we're talking about investing and spending and the financial parameters for the business. On the right side of the page is sort of a fun graph -- more than a year ago, when we were having a conversation, we were talking about a $1 trillion semiconductor industry by 2030.

Now you can see that the estimates are between $1.7 trillion and $3. Now fair enough, some of this is pricing at the device level, and you don't need more wafer starts for pricing. But there's plenty of wafer starts required when you saw some of the system requirements at the package level that showed all those devices that have to go into AI systems. -- fabs are being built at a faster and faster pace. And when we look on the left side of the page, it's not just the AI and the semiconductor devices.

All of the arrows are indicators that we follow, all of the arrows are pointing in the same direction. Hyperscaler CapEx over $1 trillion, token consumption accelerating the use of AI agents accelerating the cost of those coming down package sizes, the number of transistors on a chip, the number of chips in a package, accelerating and making that package business grow strongly. Fab utilization, leading-edge logic, 100% and DRAM, 100%; NAND, almost 100%. ICAPs are more mature process node, customer processes healthy zone now after a couple of years of digestion. It's a favorable environment for WFE to be sure. We talked about how leading edge logic, DRAM and advanced packaging will be the engines of the growth for the next few years. We talked about our strong position in that.

Those markets will be responsible for approximately 75% of the WFE growth looking forward in this forecast range. And with our leadership, we expect to grow our market share more than 2 percentage points. Finally, Gary highlighted, it won't probably be a smooth curve of growth people are talking about in studying data center energy availability. We've said that clean room capacity will be a key determinant of how much equipment can be consumed in any particular quarter. And then supply chain. We've raised our forecast on our supply chain every single quarter this year. Everybody is racing to add output.

And those things will determine how smooth the ride is as we move forward to serve this demand function. All right. Thinking about the global investments that we're making. The map gives you an indication. First of all, Epic was our largest investment. Some of you saw it yesterday, but we invest strongly in the United States. So the EPIC center is a headline investment. We're also investing in Arizona, Massachusetts in Texas, Texas where our largest manufacturing site is -- so manufacturing labs, research center, logistics centers, all of these in the U.S.

In Italy, in Israel, new manufacturing capacity, new lab space, training center, new manufacturing capacity and R&D labs in India, R&D offices, lab infrastructure and supply chain investments, Singapore, we just celebrated the opening of our second manufacturing facility there and the expansion of our capabilities. That's helping us serve almost the entire Asia customer base from our Singapore operations. And then in South Korea and Taiwan, more customer-facing investments. So a significant amount of reinvestment in the business. We're investing to grow our capacity from a manufacturing perspective and a lab perspective, more than 2x across this horizon, EPIC is a good example of that on the lab side and diversifying our supply chain globally over this time.

I don't have a specific target for CapEx, but CapEx was approximately 8% of revenue this year, and we say it will decline as a percentage of revenue as we move through the forecast horizon. Okay. AI. Fantastic results at Applied Materials. We put a few examples here to share with you. So the first 1 in our hardware development, our actual product development we've piloted the use of AI in terms of simulation and helping speed up the actual process and achieved a more than 4x on the pilots that we've run along these lines. We're implementing that on 20 products now, and we'll expect AI and the types of models that help you do the hardware development to massively accelerate the pace of hardware development going forward.

On the software side, software development, 100% of our software developers are using the AI method. So faster coding, testing, release cycles, this is amazing to get to 100%. We report the average productivity gain of 44%. But I'll tell you some of the individual projects we see are 200, 300, 400 improvement -- so 44% is across the entire company, just amazing. And then from an adoption perspective, we have more than 80% of our employees using it every single week, token consumption rising fairly dramatically, 45 in 1 year. We feel great. We think it's investable. We not surprised the industry is investing. We're not surprised that the pace of AI data centers required, we're not surprised that any of this based on our experience. And now I'd like to play a short video from Dr. Yu Cao, our Chief AI Officer at the company. Please queue the video.

I am Yu Chao, Chief AI Officer at Applied Materials. Applied created this role because AI isn't a side project for us. It's central to how we design products run processes and serve our customers. Applied sits on both sides of this market. We are a critical supplier to the AI build-out, and we are also a serious enterprise buyer and adopter of AI. Here's what we see inside applied. Every software engineer now writes production code with AI with very substantial productivity gain. And across various organizations, engineers report multiple hours of time savings per week, and they are reinvesting that time into harder problems, not easier ones.

Over 80% of our employees use AI tools every single week. We spent real money on this because the return is real. It's a P&L decision made by our management team and rejustify every quarter. As models get cheaper and more efficient, our spending goes up, not down. A lower cost per token doesn't shrink the demand. It unlocks work that simply wasn't economic before. We see efficiency as an expansion mechanism, not a ceiling of spending. For applied the prices acceleration, AI-enabled design is taken years out of our product development -- we're also targeting multiyear compression of the materials discovery cycle. That is a value creation story first, faster time to market, more share and faster time to solution for our customers. with cost savings as a secondary benefit. We are early in the AI revolution. The opportunity ahead of us is larger than what's behind us. And every layer of it starts with materials engineering. Thank you.

Excellent. Thank you, Dr. Chao. So thinking about AI and operating leverage, I shared how we're using AI in some of our product development, software development, et cetera, and seeing fantastic opportunities. When we think about our operating spending for the company, first thing to share for the investors is our operating spending as a share of a percent of revenue has dropped from 22% at the beginning of the 10-year horizon to 17% today. You can see the R&D curve I mentioned before, it was growing to -- growing to $4.2 billion, 72% of our spending.

So when we think about adding resources in the company, what sort of investments we're making, most of our additions are focused on R&D and the portfolio becomes -- of spend becomes more and more tuned to R&D as we're going through. In fact, you can see that SG&A, we bent the curve a little bit in the past couple of years. So revenue is growing faster than R&D, but we're continuing to invest. We gave you the focus, which is the most enabling portfolio. We expect that on SG&A, with the adoption of AI and the adoption of automation at a faster pace going forward that will get significant efficiencies from our corporate functions and from SG&A in general, which will allow us to continue to grow profitability and operating leverage.

Next, I'll cover our 2030 calendar year targets. So the first 1 on the top left, we've talked about 2 percentage points of additional share. So we're not giving a WFE forecast. And the reason is, of course, that's changing at a rapid rate -- what we are giving you is how we think about it internally and what our own framework is. We expect to outgrow WFE. Our target is to outgrow WFE. We think more than 2 percentage points of share will gain across this horizon. And again, you heard the reasons leading edge logic, DRAM, advanced packaging, as those -- we are gaining share today and as those continue to evolve in technology, like to foresee like the 3D DRAM, like to CFE like to panel packaging, hybrid bonding, we will be gaining share with those transitions as the industry moves forward. So it's not a WFE model, but you can model our equipment business based on growing faster than the industry.

For the Applied Services business, Tim talked about a 15 mid-teen CAGR growth rate -- and we modeled that as greater than $12 billion by 2030 with that type of growth rate. And that business, Tim highlighted how it's an excellent source of recurring revenue for the company. Other revenue in our other category, Gary headlined 2 businesses, our display business and the MAX OLED investment that we have that we're looking forward to significant adoption and growth and then the SENS Waveguide photonics business that Gary talked about, we expect those businesses to be more than $2 billion by 2030. 4% non-GAAP gross margin is our target for 2030 and 40% non-GAAP operating margin, representing expanding operating leverage beyond the margin improvement.

The bottom left, you see our tax rate, we have a 15% estimated tax rate for 2030. And those should give you the elements of our own targets how we're thinking about the company and allow you to size the company as the market expectations continue to change for WFE. From a capital allocation priority perspective, hopefully, no surprises here. First, just to reiterate the message, we're reinvesting more than 50% of our cash generated back into the business. And I gave you a good perspective on CapEx, R&D, et cetera. We'll maintain -- continue to maintain a strong and flexible investment-grade balance sheet.

And then the bottom bullet, we're committed to 80% to 100% return distributions of our free cash flow over time. And you may have questions about what's -- what will be happening with the dividend or what will be happening with buybacks. On the dividend, our Board will make that decision next year per normal. But our history here is since 2005 when the dividend was initiated, it's a 14% CAGR. We doubled the dividend per share over the last 4 years.

So a significant focus there. And then from a buyback perspective, over the last 10 years, we decreased our shares outstanding by more than 50%. So again, committed to 80% to 100% distribution of free cash flow going forward. And then just a summary of key messages. So you heard about how we think about growing our R&D, how we focus on what we add to the portfolio. It's about being the most enabling partner for our customers. We delivered 10 years of excellent profitability, executing this enabling model. AI demand is accelerating the whole outlook and certainly the outlook for our business. we're doubling. We're investing to double our lab capacity and double our manufacturing capacity from even today, so higher than the last time I said that.

And we're going to continue with our targets to grow our profitability, grow our shareholder returns, continue reinvesting in the business and continue to focus on being the most enabling company. Thank you for your attention. I appreciate everybody attending. And I'll bring -- ask Mike Sullivan to come back up and take us to the next steps. Thank you.

All right. Well, thank you so much for hanging with us here. We're going to bring all of the speakers back out and do a Q&A session. We know you have a lot to do at Semicon West. So we promised to have you out of here by 100 -- and if we run at -- if you still have a question when we're done, we'll ask the presenters to stay here, and we can always help you then if you still have a few minutes extra. So we'll bring out some chairs, we'll bring out the group.

Our thought here is we do have, of course, a webcast audience. And what we'd like to do is make sure that we have microphones available. And please don't start to ask a question until the mic is there because our audience won't know what you asked, and we want them to hear it. So we have Mike Runner, Cindy, I believe, on this side. We have Sonal on this side and Marty, my doubleganger is in the middle aisle. And if you have a question, go ahead and raise your hands and let Marty understand and the team understand that you have one, and he will do his best to come out.

So please, everybody come on out. join us right. Great. So we have a question right here.

Great. Thank you ever Karia from Bank of America Securities. Thanks for a very informative presentation. Gary, I was hoping to get your perspective on kind of 2 near-term industry debates. One is there's recent news about China perhaps getting to 4x square memory. I was wondering what you have seen on the ground, what is your perspective?

And if they do get there, what does that mean for WFE share of wallet for the Western toolmakers -- and the second 1 is this whole debate around memory despecking, right? What does that do? How much of that is real? And what does that do to the overall TAM of memory going forward if there is indeed despecking in some of the new -- so just kind of China and spec.

Yes. I think relative to compute memory and AI. And again, we had the EPYC events over the last couple of days -- and there's just so much innovation that's happening in compute memory. How you also connect it to logic. Again, we were there with all of the leaders that were driving AI server architectures going forward. So I just believe that that pace of innovation is very, very fast. And I believe the companies, the 3 companies that are leading today will continue to lead into the future because, again, there's just so much optimization, those leading companies that are designing those new architectures, what they said when we had that event yesterday morning, and they -- all those people were there, is that they have already very deep partnerships for compute memory for logic innovation, but they need to even go faster because those workloads are changing -- so I have very high confidence there's going to be tremendous growth rates still in compute memory DRAM, and there's going to be tremendous innovation at the component level, in the packaging level and the system level and then the lease backing, Bryce.

Yes, we looked at despecking and our conclusion at this point is DRAM, we expect to still be constrained in what we think the system architects are doing is they're determining that you need less DRAM for some of the different AI workloads. So training was a very heavy different amounts for inference, different amounts for Agentic.

So we think what's happening is they're tuning those systems to maximize the amount of output that you can get from a compute perspective for those different workloads optimizing the amount of memory that's used on each of those systems. So that's our perspective. We expect DRAM will remain constrained for a period of time.

Yes. That's right. And then having seen the analysis, the other thing to bear in mind is in the training workloads, they still want as many as tall as stacks as they can get their hands on. The new workloads that may not need as much as that because they can't drive utilization all the way up to the top of the HPM staff, those incremental workloads they are new, and they're additive to the market. And so it's -- everything is increasing. It's not a cannibalization. It's an increase in demand in an area where you may -- instead of having your IO and your storage in the same space, put your storage elsewhere in the system. So it's just a repartitioning -- all right.

And then Marty? Yes. Great. .

CJ Muse with Kenner for Strand. First question, Gary or Peru, Much of your discussion on the leading edge was focused on compute and DRAM. Curious if you could speak to how you're attacking optical networking from pluggables to all the way to CPO, wave guide hybrid bonding, new materials. Can you speak to that? And then Bryce, our gross margins only up 400 bps in this 4-year time frame. I would think mix-wise, just from silicon you're up 100 to 200 bps. So we'd love to hear what would drive the incremental? And are you being conservative? .

So I'll start, C.J. Relative to Photonics, we talked about that being a very fast-growing business for us. Photonics and power. There's tremendous innovations that we're driving there. In hybrid bonding, is a very important innovation, more and more vertical architectures, more bonding, better performance, better power and better thermals. I think on some of these architectures, it's not just the better performance, tokens per second per watt, but also yield. And so we're driving -- we have deep partnerships with all of these leading customers. and we're making tremendous progress on the yields. But that really, CJ, is super important on when those inflections happen. So then the second part of the question was... .

Yes. So then we had another question on gross margin for price. .

Yes, sure. So 54% is our target company gross margin. Of course, I remind investors, we're over 55% on the equipment itself today. So 54%, approximately 100 basis points each year, although we're not calling it to be a smooth ride like we talked about. We do have to invest in our infrastructure. I talked about new factory. I talked about new capabilities from a logistics perspective. There's headwinds to manage.

So we think that with fuel costs and the like, rare earths and those sorts of things. So we feel that 54% is a good target. It will require us to be continuing to launch new equipment that's more enabling we are raising pricing as part of the process, which I shared to address some of the costs that we're seeing. So CJ, we think that's a good target.

It's Tim Arcuri at UBS. I think this is all great. And I could see how a bar, though, would say they would look at the WFE share of gaining 2100 bps through 2030, and they would look at 2025, and you lost that much share in 2025, and your share in 2024 was down actually off of 2023. So the bear would say, "Well, you're only getting back to where you were in 2024 and 2025. So how would you address that?"

Well, I'll start and maybe somebody would add -- of course, in our math, we're still modeling growth in China and growth in restricted accounts in China. We can't serve leading edge. We can't serve DRAM. We can't serve NAND in China, we're modeling growth there. So you're right, if we zeroed in on all the accounts that we can serve, we'll have a greater share gain than that. And if you look at some of the device end markets, we expect a greater share gain. But if you account for the growth that we're expecting in the restricted markets in China, and again, we're assuming no change in the trade rules, then we think that's a good target for us. .

And I think an interesting point is if you do the math on it, if you look at what happened to us from a trade restriction perspective in 2025, we had access in the early 2020s to 100% of China's WFE market. then the access dropped to 90%. In 2025, it went from 90 to closer to 70, okay?

That was a big step down. And there hasn't been a change in trade restrictions that will allow us to go back up to where we were. So if you do the math on this, if you look at our position and you took that factor of trade restrictions out of the equation and then looked at our share, we're actually up and we'll show that to you off-line if you'd like to see it. So we're doing very well where we can compete even in China, where we can compete at 28 nanometers, we're doing very well. And what Prabu also showed you as he showed you, he dissected the growth rates of some of the markets, and he showed you the growth rate of leading-edge foundry logic over a very long-term period and DRAM over a long-term period and showed how we are outgrowing in those periods. So where we can compete, we're doing very well, and we expect to continue to

Atif Malik from Citigroup. I have a question for Keith. Keith, you talked about launching new optical products in first half of 2017. I was wondering if you can pull the curtain a little bit and tell us more about HBM or CVAs. The market itself is very competitive with Nova and N2 and KLA. So what will be the differentiation that LaPrairie bringing in this advanced packaging optical market? .

Right. I think it's a very important point that especially in advanced packaging, the optical inspection market is already crowded. So we would not enter a market unless we understood architecturally how we can differentiate -- and I can't really go into the details. But rest assured, we spent approximately 1.5 years examining what is serving that market today, what the technologies are in play there. And we feel that we're going to bring something to market that is clearly differentiated, clearly sustainable in its overall performance.

So we are trying to position the product to have the highest sensitivity at the highest throughput possible using cutting-edge technology. And just 1 other comment there. I think many of those products had been very much restricted in the amount of cost people are willing to put into them because historically, that market has been very cost sensitive. We don't see it as cost-sensitive any longer, just as I talked about today, there's so much money to be made in advanced packaging. The yields are so important that now we're starting to see an inflection where our customers are willing to accept higher ASPs in order to meet the yield demands. .

You guys Stacy Rasgon of Bernstein. I actually have a question for Tim. So Tim, you said you're targeting mid-teen services growth over the next 5 years. I'm a little confused. You said you thought your installed base would go from 37,000 tools to 90, which is almost a 20% CAGR just on the tool growth. And then you layer on, I guess, you said looking for 20%, I'm assuming it's not a CAGR, but 20% overall increase in revenue per tool on top of that. How do you only grow at mid-teens for the services? Shouldn't it be growing like a lot more than that if you hit those goals?

Yes. So the model is, as we mentioned, going from low double-digit growth to this mid-teens on a go-forward basis. And I think it's important to understand when you think about connected -- we're talking about connected chambers, Stacy not connected tools. So to 90,000 in the future. And as tools go out into the market, they first go into an installation and warranty period, and then after a year, 18 months, whatever it is on a customer-specific case, we moved some of those tools into service depending on what help the customers need and what sort of ROI we're able to generate from our AIX based services. .

So we're pretty confident in the growth rates that we have. The 20% that we see growing in the revenue per tool is really driven by these capabilities that we're introducing that are these AIX related examples that I gave you. But when you look at the growth in the overall market, it's really dependent upon how much of it goes to various levels of our service portfolio. We have on-demand customers who buy spares from us. We have customers who do forecast parts management. I can't go into all the details today, total kit management, and then we move into the service portfolio where we add these AIX capabilities.

So it's not a 1 for 1 as we do the transfer. But again, we see the market growing, and we see our CAGR increasing over time. And as Bryce mentioned in the last financial release, -- we did grow -- we are going to grow 20% this year based on the UT recovery as well as the adoption of the AIX capabilities. .

Yes. And if I could add a point there. I think we've disclosed at various points or installed base, which is somewhere in the high 50,000, that's tools. And so Stacy, the way to think about the growth when we talk about mid-teens is a 5% to 10% growth per year in the installed base and the balance of growth would be on new products that Tim's team delivers. .

Mark Lipacis, Evercore ISI. Thank you very much for the great presentation, really informative. I think yesterday, 1 of the presenters made the case that with a genic AI, you could have unlimited demand for compute. And you have the charts going exponential on industry revenues -- what is the -- and I completely appreciate you don't want to put a WFE number out there. I think it is a moving target. What limits the demand for your products? Is it fab capacity for your customers? Is it your customers' ability to build customers to build data centers? Is it the power grid? Like what is limiting demand right now? .

I think the primary constraint mark that we called out is clean room. So the first thing that we look at when we're forecasting medium term, next couple of years, we talk about an 8-quarter forecast that we get from our largest customers. That 8-quarter forecast is which facility, how many wafer starts a month capacity, what node, specifically what tools we have that sort of information from our customers. And so as they've worked to increase the amount of clean room that's available even in the shortest term, I think that's the major -- we talked about how that will meter the growth rate -- that is the primary variable that we see.

And then 1 thing maybe I highlighted maybe I didn't. Over the last 2 quarters, we track over 100 fab projects globally and over the last 2 quarters, we've seen more than 10 added -- 10 new projects added just in each of those 2 quarters. So the envelope of construction and the envelope of factory space and clean room becoming available over the horizon is growing. .

Great. And the next question is at the back of the room. .

It's Krish Sankar from TD Colin. Brian, just to follow up on that earlier question, you spoke about doubling. More than doubling our manufacturing capacity. Is that something you see across the ecosystem? In other words, your peers are doing the same suppliers, their suppliers are doing the same? Or do you think that ends up becoming a bottleneck for this really impressive demand scenario you outlaid? .

It's a good question, and I haven't tracked the exact comments from our peers, but I can tell you from a supply chain perspective, we have 2,000-plus suppliers we are sharing our forecast and the estimated requirements in aggregating those component requirements with all of our suppliers. So it's a good thing to think about is, yes, it's a lift across the entire supply chain at approximately that level for all of these connected -- all of the components you need in a factory.

And by the way, that's happening too, for the data centers. In fact, you see in our market how some of the end markets are competing with data center. You can't make enough smartphones today, you can't make enough PCs. There's not enough memory to go around. So there is a competition for resources. So there are some limits as we move forward. Our perspective is there are no roadblocks. I don't think we've determined any roadblocks. There are going to be pinch points that have to be managed. And I do expect that the peers will be making similar investments.

The next question here. .

It's Chris Caso from Wolfe Research. Bryce, a question for you on the margins again. And if you could address the path to the 40% operating margin. And it seems that the majority of that is coming from the increase in gross margin. So what does that imply with regard to OpEx leverage? And if you could also follow with what you expect from cash flow for the business to do as you approach those 2030 targets. .

Yes. I think those are the 2 components of reaching a 40% operating leverage we'll expect to get 4 additional points on the gross margin, and we will expect to achieve 200 to 300 basis points in operating leverage. And I'll point back to AI automation across the company will allow us to scale the corporate functions at an extremely high productivity rate as we continue to grow.

You can picture you're selling more tools to the same customer base. A lot of our functional activities will not need to expand as rapidly as revenue or the business. And from a cash flow perspective, we've averaged more than 21% free cash flow for the business. I didn't have -- don't have a target to share for that, but that's been a long-term average. We are having to invest in the business. So I think looking at that as an indicator of where we would be. I sort of expect it to be a little bit higher, but we don't have a target to share today. Great. .

Charles Shi from Needham. Maybe a high-level question. We've been hearing applies solves most critical problems for customers. But we also hear that customers from time to time kind of shifted the semi-critical noncritical applications to competitors for reasons we don't fully understand.

Maybe 3 questions related to this topic. Number one, is this applies to understanding this is happening? And two, if yes, what's the -- what do you think the reason is for that to happen? And number three, over the next 3 years, how does the team plan to address this. To me, it's probably like a low-hanging fruit for further share gains.

Yes, I can take that, Charles. So as I mentioned earlier, for our customers, 80% of the incremental wafer fab equipment growth is in those 3 areas: leading-edge foundry logic, DRAM and advanced packaging. And again, I'm meeting all the CEOs, R&D leaders on a continuous basis. What is their focus what is consuming them is the race for design wins with the best performance. And going with chip innovation and packaging innovation, it's incredibly complicated, integrating across those flows, new transistors, new wiring, new packaging architectures.

Everybody wants multi-chip connectivity with much bigger body sizes, more reticle fields, connecting together at higher performance, higher power, better thermals, so again, applied, as we mentioned earlier, we have the leadership in creating the materials. We have the leadership in materials modification, the highest resolution materials analysis -- so we're in these deep relationships cocreating architectures with all of those leading companies.

And the other thing is, as I said, more so than any point in my career they are incredibly focused on more chips out, more good chips per wafer, more wafers out, and that's the yield innovation strategy that we've been driving. So -- for me, when I'm meeting those customers, we don't have a lot of conversations, honestly, on cost right now. It's really that race for better performance and higher yields, and we're in deep strategic relationships where we're co-innovating across all of those flows for the transistor and the wiring and the high-speed memory with CMOS and logic coming to have the data going on and off the chip faster or advanced packaging.

So for them, the gains with what you had referred to on noncritical is honestly, insignificant in comparison to design wins and the engagements that we have with them. Now we are innovating across our whole portfolio. We're always innovating on every piece of equipment. We've got some dynamite innovations that we're bringing to market. But again, that's not the primary focus for our customers today.

Great. And then I think we have the next question here.

Brian Chin from Stifel. I appreciate the presentation this morning. One of the things that was outlined in the slides was are 1 of the factors that could be caused nonlinearity or non smoothing in terms of growth was supply chain. And you've obviously asked suppliers to increase double digits sequentially and then double digit sequentially again. And then again, and so when you think about that as sort of a short-term factor an inhibitor to kind of the ultimate growth trajectory, is that something that maybe we would kind of resolve in the earlier time frame of sort of the 4-, 5-year horizon -- and sort of -- there's obviously a lot of muscle memory in terms of suppliers knowing that there has been volatility over the history of the industry.

Can you inject them with money, with cash in terms of helping them along that process of scaling our capacity at levels that probably weren't prepared to do in the past.

Yes. Thanks for the question. I think the good news here is we haven't got any responses from suppliers that say they don't believe us. Everybody's seen the AI. They're seeing the use. I showed all the arrows pointing in the same direction. They see the strong demand signal. They know they have to make investments. They know they have to hire people. They know they have to expand. And so there's no conversations about sent me 5 more forecasts and convince me this is happening.

So that's a good thing. And to your point, I think -- we have studied the entire supply chain. We don't think there's road blocks, meaning that we can't get to the higher levels of output, there are pinch points that have to be managed, and there's plenty of suppliers that -- it takes lead time to hire people or it takes lead time to add fixtures or manufacturing capabilities, test capabilities, a lot of times come up. So we're asking we did raise expectations on supply chain every quarter. We're asking them to go as fast as they can, and that's what's happening.

So again, yes, I think not a smooth ride is probably a good expectation, but we don't see any roadblocks over the medium term as you called it. .

Sorry, I missed part of the conversation, but I wanted to double click more on the pricing side of things. You guys have always talked about a value-based pricing model, and that's been something you've been working on. Can you talk about in this environment? Has any aspect of your approach to pricing changed any acceleration? Just any more color on how you guys are going into those conversations? .

Yes. Thanks for the question. I don't know if anything really has changed other than some of the solutions have become a lot more valuable. And you can see that in the value that our customers are getting from their products. And so as we study the value of our solutions, let's say, it's a higher-yield solution or a higher output solution that our customers will get we can recognize the value that, that might bring to a customer, and that's a factor in the pricing.

Otherwise, I would say the mechanism itself and the discipline we're putting around for introducing a more enabling solution -- we know what it does for the process for the customer and what its value proposition is. We put an appropriate price on that. I don't think that process is changing. We're just strengthening and continuing to work on making sure we get the right price on each solution. But probably the recognition that the whole value in the market is rising, that's a factor in how we price the tools.

Great. We have a question here. .

Thanks for the follow-up. So last year, the semi industry was about $1 trillion as we go out in time. And let's assume that it grows at this is then between the different food groups, right, whether it's litho or etch or dep or other parts, which grow faster than the others.

So I'm not asking for a WFE number. I'm just saying in what scenario does intensity become lower than what it is? And then if it grows at this pace, what are the different components? How fast do they grow relatively?

Well, I'll make a brief comment and maybe probably Gary, you guys can talk about what grows faster from a process perspective. I think the fact that the price of the end device has been what's moved that equation the most. So I think if you said, well, what can I make a density go a lot higher in the short term or what can make it go even further down.

I think it will be the price of the memory or the price of the GPUs that would change that equation radically. Underneath, I think we shared how the process technologies, of course, are adding more steps. The packaging is adding more steps, so it's becoming more intense. Our business is growing. So we don't use intensity as as a factor when we're forecasting our own growth, we're back to what are the factories? What are the end markets? How fast are they growing? That's what we look at.

So I think we showed earlier those charts on the node over node changes for leading-edge foundry logic, DRAM. And again, what you see there is WFE dollars per wafer start going up significantly, our served market going up and our share of served market going up. And those architectures are more 3D architecture, 6 square with CMOS bonded to the array, 4 square, 3D DRAM, there are more material innovation enabled Same thing in foundry logic as you go to those subsequent architectures. And then 1 of the biggest innovations in the industry is multi-chip connectivity, where you're connecting computing components together with new architectures, that's very much enabled by materials innovation.

So from Applied's perspective, as we see those road maps going forward, and again, we're engaged with all of these leaders 10 years out in the future, 4 technology nodes out in the future. So we have very high clarity on what those architectures look like, and they are more materials enabled, which is a great tailwind for Applied Materials.

Yes. Maybe I can add 1 thing that -- if you really look at those, we changed our numbers and a few of them if you look at today. We increased the SAM number in certain areas. As the technology getting more and more complex, Okay. A lot more steps are ordered. You don't even see it. Some of the steps come and go. The customer keep it as an IP, a lot of steps in treatment lot of steps and Aneel. A lot of steps are being ordered. So that's the reason we increase our number today in some of the SAM numbers.

Every time I see it, I think I'm wrong because that's my SAM number is going but more steps are ordered and also in packaging, is really look at all about thermals, the thermal interface material, how do you stack out all the heat. A lot of new materials are coming in. So we are always surprised with how many more new steps are added at this stage. -- the complexity increases.

Great. All right. Well, thank you. Well, to the audience, I just wanted to let you know that I hope you don't mind that we did things a little differently this year, a lot more speakers, a lot more content, but we thought it was a good time to do that. We've been very excited to have the opportunity to present to you today. I'd like to thank all of our colleagues and their teams for the work that they put into this. And most of all, we'd like to thank you for coming and spending time with us and for your continued interest in Applied Materials.

Okay. Welcome. Good afternoon, everybody. Welcome to the Goldman Sachs Communacopia and Technology Conference. My name is Jim Schneider. I'm the semiconductor analyst here at Goldman Sachs. It's my pleasure to welcome Applied Materials and CEO, Gary Dickerson, to the stage today. Welcome, Gary.

Thank you, Jim. Glad to be here.

I think it's fair to say that the view of the market is pretty starkly different than we were on your stage last year at this time. Clearly, we see almost everything being constrained around AI. You commented on much stronger demand visibility that you're seeing from your customers. Can you maybe help frame the level of demand and visibility you're seeing right now, either in terms of duration of customer forecast or size of your backlog?

Yes. As you said, AI, especially data center, is an enormous driver for our business. One statistic I thought was interesting. If you look at leading-edge foundry, you go back a few years ago, smartphone wafer starts were 4x larger than data center. Today, they're 1:1. And in a couple of years, it will be 2:1 data center versus smartphone. So for sure, AI and data center are by far the biggest drivers. And relative to the technologies that are critical for AI data center, the fastest-growing segments, 80% of wafer fab equipment growth this year is leading foundry-logic, DRAM, including high-bandwidth memory and advanced packaging. Again, that's 80% of the wafer fab equipment growth this year, and we see that same mix in '27 and going forward.

And then relative to our customers, we've been continually increasing our revenue forecast for '26. In February, we said greater than 20% revenue growth. In May, we said greater than 30% revenue growth. Now we're approaching 40% revenue growth and what we've communicated here recently. And both our near-term and longer-term outlook has been growing.

One of the big drivers this year is Agentic AI. So you have agents consuming tokens and compute. And that's driving increase in our longer-term WFE forecast, especially around CPUs and DRAM. So we've seen that continual increase. And then relative to all of our discussions with customers, Applied is in a great position because all of those fastest-growing segments of the market are areas where Applied has clear leadership, and we're gaining share through those inflections. So customers want to make sure that from an operations, supply chain capability standpoint, we're ready. Our trained service engineers are ready. So we do have this 8-quarter forecast, rolling 8-quarter forecast and longer-term commitments from our customers. And again, that has continually went up as we progress through the years. And we even have discussions with customers all the way through 2030.

My view is AI compute demand is going to remain very strong for many years going forward into the future. I look at Applied Materials and the way we're using AI in innovation, product development, operations supply chain. We have thousands of engineers in our field operations, service and process engineers. And it's just a game changer. And then for me, what is most important, we will drive tremendous productivity improvements inside Applied Materials, but way more important than that is our ability to bring multibillion-dollar products to market faster and also very valuable service innovations to market faster.

So I think, again, AI is the biggest technology inflection I've ever seen in my life. If you look going forward with Agentic, physical AI, again, you're going to see increased compute demand and the need for compute innovation, tokens per second per watt has also never been greater. The position for Applied enabling these inflections is also a great tailwind for the company.

Great. Gary, you've seen a lot of cycles over your career for this industry. How does the current demand environment compare with past circle cycles you've seen over your career? And then is there any sort of warning sign at all that rhymes with past sort of cycle peaks that you've seen?

Yes. So I think it's pervasiveness of demand. I mean I look through my career, I'm an engineer, been innovating at 3 different companies in the industry. I've never seen a driver that's as broad and pervasive as what we're seeing today. I mean, certainly, when we went to mobile social media, everybody has a computer and a camera in their pocket. That increased the demand but AI is a whole different level of demand driver.

So when we look out for a number of years, again, it's just so disruptive and fundamental to every single industry. My view is that companies that are not adopting this technology will be left behind. And so -- and the growth, Agentic is adding additional WFE growth going forward. Physical AI in the future will add more compute demand going forward. And I don't really spend time worrying about macro effects over a near term. Again, for me, it's very clear this compute demand is going to be strong secular growth over many years going forward.

Yes. Now your direct customers are incredibly profitable right now, as profitable as I've ever seen in my career, certainly. But some of your customers' customers who are driving ultimate spending on chips, they're spending well over $1 trillion run rate at this point. That appears to be moving higher. But many of those same companies are tapping capital markets just to sustain that low current level of CapEx.

So how do you think about the ultimate returns for the AI value chain as you see it big picture? And do you see any kind of systematic risk that could kind of cause a pullback in demand?

So I think the value creation to me is crystal clear, as I mentioned, for Applied. We're bringing to market products that are worth billions of dollars. So that time to market is extremely high ROI. Another area that I would say that we're getting tremendous pull higher than any point in my career is not just for performance innovation. Again, Applied is the leader in material innovation in all of these fastest-growing parts of the market being driven by AI, but the other -- and that's critical for our customers for design wins. So we're co-innovating with them on all those architectures for generations out in the future.

The other big driver larger than I've seen is in yield and output innovation. So our customers, as you can imagine, every incremental chip that they can ship is worth an enormous amount of money for those customers. So if we can innovate in our services, and that business is growing over 20% today year-over-year. Margins are up 180 basis points in that part of our business. And so innovations that we're driving there for predictive models for preventative maintenance for more output or chamber matching where I can get more good chips output per wafer. Those are enormously valuable for our customers.

So again, for me, what I -- what we're focused on are driving products to market faster that are worth billions of dollars, and we're doing that. It's very, very clear. Same thing with these very high-value services. That 20% growth in services over the last year, again, we see the opportunity to create more valuable services for our customers because the value of the yield and output is so high. In some cases, if we go to a different service contract, we can more than double revenue per tool for those innovations. So again, for me, that is a complete game changer in terms of Applied's position in the industry, the value that we create for our customers. And again, I have to believe that everybody else is looking at this in a similar way.

So as I said earlier, if you're not implementing these technologies, you will absolutely -- by the way, I hope our competitors are not doing this. They're going to be left behind. There's no question. And so relative to financing or any of those types of dynamics, what I look at more so are the fundamental value creation drivers. As you said, the profit pools in this industry are larger than they've ever been because this technology is more disruptive than any technology that we've delivered in the past. And so I think this industry, we create miracles, if you look at the technologies that we've been able to deliver in the past that have had enormous impact. This is a whole another level.

So frankly, I don't worry so much about what you just talked about. I think that this is so fundamental, the value creation is so significant that, that compute demand is going to increase significantly going forward. Now are there going to be periods of time where it's not growing as fast for whatever type of a macro effect? That could happen. But I think the trend over a number of years to me is crystal clear.

Yes. Now as you said, AI permeates everything we've talked about at this conference. One of the things I'd like to do is explore how companies are looking at AI internally. So you just raised that question. So maybe give us examples of how Applied is using AI internally. And are those mostly kind of cost efficiency and efficiency-led kind of applications or using it to drive higher revenue for the company in some cases?

By far and away. So if I look at doubling the size of Applied Materials over the next few years, we definitely -- we are incredibly clear in all of the swim lanes that we're driving relative to AI implementation in the categories I talked about earlier. By far and away, most important to me is top line growth, bringing products to market that are worth billions of dollars. One example is in coding. In some of our products, there's a high software and algorithm content. And we're driving tremendous productivity in that part of the product development. So 50% improvement in productivity has nothing to do really near term with headcount. It's all about bringing those multibillion-dollar products to market at a higher velocity. So we're using this technology in a number of different areas, hardware and software. There's no question that the value is tremendous. And I mentioned the value of yield and output innovation and services.

So when I'm meeting some of our top CEOs every month, they're racing for tokens per second per watt in their new chip and packaging architectures, that's really important. But they're also phenomenally focused on increasing output of good chips per wafer and wafers through the use of these kinds of technologies. In our services, we have over 37,000 chambers in the field today connected to AI-enabled servers, most of them connected remotely. And so again, our ability to -- there's a massive amount of data there. Our ability to use AI with predictive models to increase output and good chips per wafer is off the charts valuable. And as I said, that business is growing -- has grown in the last year over 20%. The value there is higher than ever. And in some cases, we're improving the hours per task in order of magnitude. But for me, what I care about is I can accelerate scaling these high-value yield and output services faster for top line growth. So we will add headcount, but we're going to grow revenue many times faster.

Understand. Okay. I want to ask just briefly about U.S. semiconductor capacity because obviously, the U.S. has lagged the world in the last 20 years plus. But it seems like that is starting to change a little bit with increased investments from TSMC in Arizona, Samsung, potentially Intel. So maybe speak to your market position, specifically in the U.S. market. Do you think it basically matches what you see globally? Or is it higher or lower? Maybe talk about why.

Yes. So I think relative to market share, the most important thing for us is how we're positioned in the technology inflections. So Applied is the leader in leading edge foundry. We're working across all of these different companies. We have the most valuable enabling portfolio, whether it's in transistor inflections, backside power, all of those different areas.

In DRAM, one of the big inflections there is in higher speed DRAM. So adding transistor innovations in the periphery to get the data on and off the chip, that's an area where Applied has extremely high market share. And again, that's another inflection that's important for our customers. And then packaging, multichip connectivity, that is also one of the fastest-growing segments of the market -- overall semiconductor market. That's growing over 70% this year for Applied Materials, kind of midpoint between $2.5 billion or $2 billion and $3 billion relative to the size of that business, extremely high growth rate.

So for us, it's really more of our positions and enabling those really important AI data center inflections than it is whether it's U.S.-based or Asia or any other location. Now the one thing I would also say, though, that is a very big positive for us, as customers are moving into new locations, that is a tailwind for our service business because that infrastructure is not set up in those new regions. And also the level of experience to ramp yield very quickly is less in some of those new regions. So I would say that we're well positioned in terms of those technology inflections. And the regionalization is a tailwind for us in our service business.

Yes. One specific I want to ask -- example I want to ask about is Elon Musk and who's being more vocal about building Terrafab in the U.S. Can you talk a little bit about the extent of your technical and commercial collaboration with Terrafab, sort of what you expect your market position to be with Terrafab as it ramps? And then maybe how big an opportunity this could be for you over time?

Yes. So I'm not going to talk about -- I have a lot of respect for Elon, very smart person. I'm not going to talk about any specific customer. What I would say is that we're engaged with every one of these companies. They're all in a race for AI compute leadership. So every technology node, whether it's 2-nanometer or A14, 1.4 or different DRAM technologies, I said, adding the transistor innovation in the periphery for higher speed memory, advanced packaging, multichip connectivity. We're engaged with every company. Applied has the broadest, most unique portfolio in enabling those technical innovations. Applied is by far the leader in deposition for all of those different segments and essential for customers in bringing those new innovations to market.

So we're in deep high velocities, co-innovation relationships, co-creating those architectures, whether it's a new customer or an existing customer, the depth of the relationships are very strong. When you have a new customer, the most important first thing is the R&D relationship to create competitive technology. So we -- every single company, we have those deep relationships in all of those different areas. And then how that transitions into high-volume manufacturing depends on how fast you can create a competitive technology.

And then it's not, as I said, just about performance. For design wins, you need performance and yield. And again, when you have new companies that are entering into these markets without that level of expertise, it creates an even much stronger pull, not just for our innovation teams, but also for our yield and output teams.

Great. And then maybe last regional question on China because this has been an area where there seems to be like a resurgence of CapEx on the part of those customers, but U.S. export controls have obviously put a damper on how much of that market you can address. You've talked about that market now being expected to be up for you in 2026. Maybe help profile for us how do you expect that level of structural growth in China for Applied and in light of all those kind of export controls you're subject to?

Yes. So China really for us is mostly an ICAPS, we call ICAPS market, IoT, communication, auto, power, sensors, more consumer electronics. And as I mentioned earlier, 80% of the WFE growth in '26, we think the profile is similar in '27 going forward is really driven by AI data center, leading-edge foundry logic, DRAM, high-bandwidth memory, advanced packaging, those segments.

For us, we're seeing growth this year in China and overall ICAPS. And gaining a little bit of share in the areas where we can compete inside of China. So we think longer term, if you think about the fastest-growing segments of WFE, again, it's still going to be driven by AI and data center, the segments I described earlier. We think ICAPS grows at kind of mid- to high single digits over time, but not as fast as some of those other segments. And what we're seeing near term, even in ICAPS, the growth there is being driven by power in data center, photonics, that's adding some incremental growth in ICAPS.

Next year, we see that business also growing a little bit faster than what we're seeing this year, but still growth rates are slower than what we see in data center. And again, one thing that I've seen over the last few years was that smartphone ratio to DRAM, 4:1 going to 1:1 and 2:1, the data center versus smartphone, that's going to be the driver for the next several years.

Yes. Okay. So maybe shifting to a discussion of your end markets for a moment. I don't -- I know you don't like to give quantitative market forecast for WFE growth, but I'd love to get your qualitative perspective on 2027, if I could. You've seen obviously very robust demand visibility as you talked about. One constraint you've mentioned on your conference call back in August was clean room space. That was a constraint this year, talk about being constrained next year. But this year, your customers kind of overcame that and accelerated that space completion. So directionally, how would you sort of handicap the chances of growth in 2027 for you being better than '26?

So growth in '27 will be strong. Again, compute demand is increasing. I mentioned Agentic on top of what we have been forecasting previously and physical AI going forward. So '27 is going to be strong and beyond '27, growth is going to be very strong. So I'm not going to give a specific number for '27 at this point, but '27 is going to be a great year for Applied.

Got it. But you see growth into 2028 as well?

Again, I just believe compute demand, this technology is going to transform every single industry. Compute consumption is going to keep going up. Again, you have a Agentic adoption where the compute is being -- compute demand is being driven by agents in addition to people and then physical AI on top of that going forward in the future. So I just think that compute demand is going to keep going higher.

Yes. Now specifically on memory, that really appears to be leading growth in the industry, at least short term. Do you see that relative strength continuing into next year? And or is there a point in time where you expect, say, leading-edge foundry or one of the other areas to sort of to grow faster?

So the 80% is in the areas I talked about, including both of those segments you just described. I think longer term, those are going to be 2 of the fastest-growing segments. I'm not going to give exact numbers here today, but both of those areas are -- if I thought about what are the areas that are going to grow the fastest, there's no question that leading-edge foundry is going to grow at a very high rate going forward. DRAM will also grow at a very high rate going forward. In advanced packaging, the multichip connectivity is one of the biggest areas of innovation in the entire industry. That's also going to grow very quickly.

So I don't know I want to handicap one versus the other. They're both going to be in that Top tier of growth, those segments. And again, I would say ICAPS, that is not as exposed to the AI data center growth. So that will be kind of mid- to high single digits. Those other markets will be growing at a faster pace.

Now advanced packaging, you just mentioned, that's an area outgrowing the industry as well. There's expanding range of competitive offerings from different vendors there. What's your revenue currently for that segment? And what are your expectations for the ramp? And sorry, has the expectations for your ramp in that revenue stream changed at all in the last few months?

So I think how you connect computing components is enormously important for the industry. So these system companies, they would love to have the package -- the body size as large as possible, as many computing components connected as possible at the highest I/O density. So I'm highly confident in the next 3 to 5 years, you'll see incredible inflections in this technology beyond where we're at right now. This is a really important race for AI computing leadership.

So for Applied, that business has grown -- is forecasted to grow over 70% this year, kind of midpoint between $2 billion, $3 billion in terms of the size for Applied Materials. And I would say going forward, again, I would put that in top tier relative to growth rates. Applied, that's one of our strongest segments from an overall market share standpoint. High-bandwidth memory and 3D IC are the fastest-growing segments this year. And that's, again, where you're stacking chips closer together for higher performance and lower power. And there will be tremendous innovations.

Applied has this broad, unique portfolio. We just acquired a company in panel-level plating. That will be a great return for us. We acquired another company with X-ray technology that's important for advanced packaging. Applied, our inspection measurement business is growing over 50% this calendar year. Packaging is an area where Applied has leadership in all of those process equipment technologies, but also we have unique technologies for yield innovation in advanced packaging. And that will be one of the fastest-growing segments for us going forward. There's no question.

Excellent. So I want to ask you a couple of product level questions for you, too. On the earnings call last month, you talked about process control. That segment for you growing over 50% this year. What is it about that segment that kind of gets you so excited and gives you confidence in growth?

I love that segment. So again, as I mentioned earlier, what's important for customers is winning inflections in computing performance and power. So that is something that every single customer, we have deep co-innovation relationships with their teams. But also yield innovation is absolutely critical for design win, you got to have both. You got to have performance at a certain level and yield at a certain level. Applied is the leader in the industry by a large amount in electron beam technology. That business is growing at a very high pace. We have very unique technology leadership in imaging, speed of imaging. And for material analysis, this is becoming more and more and more important.

So both our e-beam technology and our optical technology are growing at a very fast pace, over 50% this year. I have very high confidence, especially with the products that we have in the pipeline that this is going to be a very high-growth business for Applied for a number of years in the future.

And one thing I would say that's also unique for Applied Materials is the combination the materials innovation in chips and packaging, Applied has strong leadership. If you think about what is enabling power and performance in a chip and package, Applied is much larger than our nearest competitor. Then we are unique in that we have this leadership in e-beam technology so we can drive not only performance innovation, but also yield innovation. And that combination is incredibly important because that also gives us visibility in terms of how to drive innovations in our process equipment at a faster pace. So that combination certainly as a growth driver, 50% growth in revenue, great growth driver. It's going to be a great growth driver going forward, but that combination also is unique for our entire portfolio.

Okay. I want to ask a couple of financial questions. But before I do that, I want to get an update from you on EPIC because that's an area of customer collaboration. Obviously, you're trying to be close to your customer to drive ultimately revenue market share. How are investors going to know that it's working?

So EPIC, again, as I mentioned, every customer, they're in a race for AI compute leadership. And it's all about time to market in bringing those new technologies to market. Innovations in advanced foundry logic. Innovations, I mentioned the CMOS transistor innovation for high-speed memory or advanced packaging innovations. And with EPIC, this is where those architectures will be created. So bringing our top customers and other ecosystem partners together to innovate, to create these new architectures is incredibly important for our customers and for Applied Materials.

For us, it enables us to be designed in with our systems and our advanced services for yield innovation as those new architectures are created and adopted in high volume. It gives us great visibility. We're co-creating these architectures for generations out in the future. So we have tremendous visibility in terms of where to invest, what it's going to give us the highest returns. And for the customers, it's all about design wins. Their ability to bring these new architectures to market that give them an advantage in tokens per second per watt and then being able to ramp faster than their competitors. So that's really the value for them. And we have so deep relationships with every one of these leading customers, deep relationships to co-create those architectures going forward.

Got it. Okay. Then one investor debate that I hear a lot is pricing and the ability for you to price the value you're creating in the market. Historically, that's been difficult because you've got a concentrated customer base. Maybe talk to -- but I think some of your customers have publicly stated that they see tool costs rising. So maybe help us understand your ability to raise price, both in the short term and the long term, the impact of that on margins.

Yes. So I think that we're in a position where the value we're creating is higher than ever in performance, and I also mentioned in yield and output. So if you look over the last year, we've increased -- our semi systems gross margin is in the mid-50s. And we've increased our gross margin about 190 basis points. I have very high confidence we're going to continue to drive that higher because the value in winning the race for these inflections for our customers is very high. And those profit pools are larger than they've ever been.

So for us, for them, it's really about this co-innovation relationship, bringing those technologies to market, winning designs based on being first to market with the best performance and power. And then we also increased our services margins 180 basis points in the last year. And there, again, the value of yield and output innovation, as I mentioned, is higher by far than I've ever seen at any point in my career. And the good news is we have those over 37,000 chambers connected to AI servers. We're bringing to market many service innovations that are incredibly valuable. We're increasing revenue per tool, and we're increasing the margins.

But you can imagine the incremental profit for our customers for every wafer and chip increased is off the charts valuable. So that's pulling those service innovations. And again, it's all about the race for AI compute leadership, and we have tremendous value that we're creating there. So I have high confidence. We've been increasing margins, high confidence we'll do that going forward.

Excellent. Closing on M&A. This is something that's kind of come in and out of the aperture of Applied over time. How are you thinking about M&A as a strategic priority? Do you see opportunities? Is it more important or less important longer term than it was before?

Yes. I think that -- so in terms -- we're going to generate a lot of cash. We'll return 80% to 100% of cash to shareholders. I wish we could do more M&A. But in the current geopolitical environment, it's just not possible to do large M&A. We have done -- I mentioned in panel plating. I think that was a great acquisition, small acquisition we did recently. The return is going to be great for that. We did another acquisition with X-ray technology. Again, that's going to be a really great acquisition for us.

So we're doing those smaller types of M&A. We've also made some strategic investments where we've had really good returns that strengthen Applied's technology opportunities. Hybrid bonding is an area where that is going to be a really important technology for compute innovation, higher I/O density and synergistic where we're combining a partner technology with 5 technologies from Applied Materials on an Applied Materials platform. So we're doing some of those strategic investments. We're investing in small M&A, but really most of the cash will be returned to shareholders.

Great. Well, I think with that, we are out of time. Gary, thanks so much for doing this. Really appreciate it.

Yes. Thank you so much. Thank you, guys.

All right. Good morning. Welcome to day 1 of Citi Global TMT Conference. My name is Atif Malik. I cover U.S. semiconductors and semiconductor equipment stocks. It's my pleasure to welcome Brice Hill, Senior Vice President and CFO from Applied Materials; also Mike Sullivan, Corporate VP and Investor Relations.

I'm going to kick it off with my questions first. If you have a question, save it for the end. I'll call the mic to come to you, and you can raise your hand and you can ask your question. Welcome, Brice.

Good morning.

Brice, let's start with the market outlook. A very exciting year. You've described customer visibility as the best you've had. Rolling 8-quarter forecast, tracking 10-plus new fabs, having conversations extending to 2030. You have been raising your semis revenue expectations to more than 30% now. Can you just talk about the confidence on this continued growth and how you're looking at your opportunity in the next few years?

Sure. Nice to see everyone. Thanks for hosting us at the conference, Atif. So the demand confidence that we have, we look at -- we have a lot of information coming in from our customers, especially the DRAM and leading logic companies. We have a rolling 8-quarter forecast we get from our largest customers, which has been increasing all year, and we attribute that to the AI, artificial intelligence systems demand that's happening across the market. So we've got that information coming in directly from our customers that's been rising. That gives us some confidence.

When we look beyond our customers to their customers, cloud service providers as an example, we see the CapEx forecast rising all year from the cloud service providers. I think it's over $700 billion this year for the U.S. companies, almost $1 trillion for the global companies. Those numbers continue to go up. When we look at the underlying trend dynamics of AI itself, we see token demand rising. We see the economics of token demand and the economics of the cloud service providers to be positive. And so this underlying trend seems very strong.

And when we think about it from a microeconomic perspective, what we see inside Applied Materials, we're, of course, using AI inside the company to accelerate software development, to accelerate product development, to accelerate schedules for our products. And our own utilization is increasing significantly, which gives us a lot of confidence that this trend, the AI data center investment trend is investable and is durable from an economic value perspective.

The last thing on this is we've seen across the whole ecosystem, we've seen utilizations increasing almost every different end market, including our mature technology nodes, which just says that the semi devices themselves are increasing in volume, which also gives us confidence. So really, all the indicators, all the arrows point up, our customers are investing, our customers' customers are investing and highly profitable, and we're investing. So we think we have a lot of confidence in the future.

Brice, one thing that we've started to do in terms of forecasting wafer fab equipment, I understand you guys really don't talk about that number is start using the hyperscaler forecast. And you guys have a good team to do this kind of bottoms-up. And kind of help us understand if you guys are doing anything differently in this spending environment, tracking like the hyperscalers' investments or any other metrics that is extending your visibility?

Well, for sure, we're monitoring actual system designs as an example. So we do look at the GPU content, the accelerator content, CPU content, memory content on systems. We're trying to understand what the shape of the demand by DRAM, by leading-edge logic and by advanced packaging will be for our forward-looking forecast. So that's probably at a deeper level than we've seen before.

The other thing, of course, that we track is all the factories announced across the world. So we have a fab tracker -- it's well over 100 factories that we're tracking at this point. There were 10 -- more than 10 added in each of the last 2 quarters. And that's another strong leading indicator for us is how much clean room is being put in place. We've said over the medium term, this ramp, there's more demand than there is supply. So we've said in the medium term, it's gated by clean room.

And so we're looking closely at the investments customers are making in clean room when those factories come on, and that gives us confidence also and then the triangulation of that system architecture and the amount of capacity that's being put in place across those different device types, we work hard on that.

Great. Let's talk about the growth drivers. You guys have talked about foundry-logic, advanced packaging and DRAM driving more than 80% of your growth this year and next year. Can you talk about the areas like NAND capacity and mature node our ICAPs business, power semiconductors, silicon photonics, -- like which one of these areas have the most potential for upside to your current already very strong expectations for DRAM and foundry-logic?

I think the -- so if we start with DRAM and foundry-logic, the end market that we would pair with that very closely is advanced packaging. And so the advanced packaging, that was a $1.4 billion business for Applied Materials last year. We've said this year, it's growing more than 70%. And we expect that advanced packaging will grow apace with leading-edge logic and DRAM going forward. And it goes back to what I just described with the system-level architecture for AI data centers.

You have a lot of leading-edge components, CPUs, GPUs and accelerators. You have stacks of DRAM and HBM on those systems, and they all have to be connected on a substrate with high-performance interconnects. And so that pulls the advanced packaging capability along with the demand for these systems. So we think that's a very strong grower. We put those 3 together and Applied has the #1 process equipment position in each of those end markets.

The other one that's a little bit different than in the past year is what we call ICAPS, our ICAPS market, the mature process technologies. ICAPS is IoT, communications, auto, power, sensors, those end markets. That's been a slow market for the last couple of years as a lot of capacity had been put in place in China. And we said the world has to digest that capacity and get utilizations back up. Well, now we see utilizations coming up. We actually expect some moderate growth in ICAPS this year in 2026. And next year, we think will be a more normal growth year for ICAPS.

Normal, meaning the end market, the devices themselves grow at mid- to high single digits, and we expect the equipment will eventually track that as we normalize from a utilization perspective. And so that's probably a change. If we have our ICAPS business working well along with DRAM leading logic and advanced packaging, then we can get the strong results that we're expecting.

On NAND, I'll just add a comment on NAND. The bit demand for NAND is very strong. So we say in the probably high 20% range. The issue is that the additional layers that are being added in the NAND process technology is so productive from adding bits per square centimeter on the wafer that you actually need fewer wafers to complete all of the demand or to make all of the demand. And so when you do that, basically, what you have is an upgrade cycle on the NAND side. So you have fewer new factories, you don't have more wafer starts, you have a lot of upgrades. So what that does from an equipment perspective is it makes that market smaller. And we expect that to continue for at least the next few years.

Great. Brice, at June, DRAM and advanced packaging masterclass, you highlighted panel-level packaging. It's something that you guys have been kind of working towards in the last few years. Could you talk about where the industry is today on panel adoption? And what are the key hurdles and opportunities for Applied Materials?

Sure. So panels, if you picture these data center systems, all of the large cloud service providers and system design companies, again, are looking at their mix of components, accelerators, GPUs, CPUs, stacks of memory. They're trying to figure out how can they get even more performance per watt, even more performance because there's such a concern about the amount of energy being needed to run those systems. And one of the most important ways of doing that is to expand the amount of chips with higher quality interconnects on a single system. And that requires larger substrates, panel being one of the options for larger substrates.

And Applied Materials is investing significantly in this trend to grow its capabilities. We just recently did a small acquisition. The company is called NEXX. And as an example, that does fine line interconnects on the on-panel levels. You ask where the industry is. The industry is currently in development on panels. We have revenue on our panel processing equipment, but I don't think there's any volume production at this point. So it's in development.

From -- what are the challenges? I think the challenges are typical of what you see in new package technologies. The engineering has to do with how do we solve the thermal issues on a panel in a different substrate, how do you maximize the density and just the form factors themselves, companies are experimenting with different form factors. Some are wafer size and some are larger than wafer size. And by the way, Applied's experience on the display business gives us the ability to experiment and drive equipment solutions in those larger form factor sizes.

So we've invested in litho capabilities. We have e-beam products that will help inspect on a panel or other package types. And as I mentioned, we're investing in other capabilities. So we think that will be a natural evolution in terms of larger compute systems over time.

Brice, one question we've been getting from clients is around the tightness in the DRAM market where some of the AI servers are going to be de-speccing memory content because of the availability of memory and the move down towards maybe 8-high or even lower stack that's up HBM. Can you talk about -- obviously, they have to do this to kind of adjust themselves to the availability of the memory market. But can you talk about any impact you see on the equipment demand?

Sure. I've heard this term de-speccing. And I think our view at Applied Materials is when customers are lowering the number of stacked HBM, high-bandwidth memory in a system, they're doing that to get more system output. And so from our perspective, they're going -- they're trying to be able to ship more AI data center systems in total, dealing with the memory constraint. And so they're adjusting the system throughput and orchestration in that system so that they can maximize the amount of compute with the hardware that they have.

What that means on the equipment side is you're actually shipping more GPUs, more CPUs, more accelerators per memory chip. So I think from an equipment perspective, we would view that overall as neutral. And we certainly haven't seen a reduction in demand in the business because of this phenomenon. So I would just look at it like any other situation where you're managing the amount of chips available, they're trying to maximize the amount of compute they're outputting with their systems. And so it's just an adjustment to figure out what's best, the most memory or the most systems, and they're leaning toward more systems in that de-speccing situation.

Great. Brice, when we talk to some of your customers like big foundries and ask them, how is this cycle any different from prior cycles? One thing to point out is that the prior cycles were all about Moore's Law, just making it smaller. But with this new age of AI, the performance and the time to get the latency and those improvements is far more stringent and they're really counting on the equipment makers to provide that extra boost in performance. And you guys have the most kind of broad co-optimized portfolio. You talked about integrated process tool set. Are customers increasingly awarding you more PTORs or process tool of records because of this ability to boost the device performance in a more integrated manner?

Great. Thank you. I think that's something that I want -- I was hoping to touch on. So that's a good way to think about Applied's portfolio. When we think about our portfolio of tools and investors know that we have a large number of solutions at Applied Materials. We have deposition, we have etch, we have treatments, we have films. We have thermals. We have inspection, we have planarization tools, et cetera. I've heard some investors say before, oh, if you invest in Applied Materials, you're just investing in the entire semi equipment stack, so the company should grow like the overall market. We certainly don't view it that way.

What we're trying to build is a complete set of processing equipment to be the most enabling set of tools for our customers. So the way we look at it as when our customers come into EPIC, our new development center that we're building into -- in California, when they come into EPIC and they want to work on nodes, process technologies that are 5 to 10 years out, next-generation technologies. And we need to invent the type of equipment, and we need to invent the technologies to build those new advanced architectures.

Applied wants to be able to do every step in that process, so they're not having to go back and forth with other process solutions. So if there's 16 steps in a particular formation of something in the architecture, we want to be able to do all 16 steps so that we can control the interaction of those materials and control the way the process is developed and optimize the tools for that solution. The benefit for our customers is they've got a one-stop shop to do that work. The benefit for Applied is when we're successful, hopefully, we win all of the designs for those applications.

So it's really a focus area for us. And so I would look at that portfolio less as a desire to be broad and more as a desire to be the most enabling technology company for our customers. And that's the way we think about recent acquisitions we've made. We've done an X-ray technology company that's going to help us inspect micro bumps between stacked chips. We did the NEXX acquisition that I mentioned that's fine layer interconnects for panels. And that's the way we think about our portfolio is what technologies will help us be the most enabling partner for the customers and be critical for our EPIC center where we're trying to invent those technologies that will be 5 to 10 years in the future.

Awesome. Now we've established the demand picture looks really strong. Let's talk about your ability to supply. You doubled your manufacturing space over the past several years, and now you're preparing 2x current systems output levels by 2028 and some planning going on to 2030. Can you just talk about when you look at your supply chain readiness to meet this demand, where you guys stand?

Sure. Well, on the capacity side, one of the most important things when you have long lead time capacity investments like clean room. Clean room typically takes 2, 3 or 4 years depending on what you're starting with. You never want to be short clean room when you're in a highly profitable business. And so when I made those comments, I wanted to make sure investors knew and our suppliers. Partially, those comments were also to be reinforcing to our suppliers who have to make the similar investments.

We have invested to -- and are investing to make sure that we can produce double the amount of systems each quarter that we're producing today. And it's a capacity statement. It's not necessarily a revenue forecast, but it tells you those are possible scenarios that we're considering. You have to have the clean room and space in place. So we're putting that in place. And when we think about our suppliers, those 8-quarter forecasts that we talk about from our large customers, we aggregate that information by component type, and we send that information to our suppliers so that they're getting 8-quarter advance notice on what type of demand that they -- that we will have of them so that they can make the investments.

Our suppliers, there's a lot of smaller suppliers. They're all different sizes. We typically reference 2,000-plus suppliers for Applied Materials. A lot of them have to make investments, hire people, add capacity just like we do. And so they need that signal. Now the only rub in the equation is we've had -- we have that 8-quarter forecast. And of course, every single quarter this year, it's risen even within the year. And so our suppliers are doing a great job keeping up with us and the customers understand the importance of sending those signals. And so we continue to work on optimizing that whole communication flow.

Great. Let's talk about gross margins, and I actually feel guilty asking you this question because you've done such a great job in expanding gross margins by 300 basis points the last few years. But you're guiding to kind of flat gross margins near term. And can you just talk about the drivers to expand these gross margins further, whether it's value-based pricing, product mix or manufacturing scale?

Sure. I think the most important thing to think about for Applied Materials gross margins and is the R&D function first. You mentioned pricing and cost and mix and those kinds of factors. What I would ask investors to think about is how we're focusing our R&D. The company, the projects that we pick with our customers, we're selecting what we think are the most enabling and the most valuable projects. And if we're successful in building those solutions with our customers, then the portfolio becomes more and more valuable every year as we introduce those new solutions and the complexity and value of those solutions becomes higher.

So pricing becomes just your mechanism to size up the value that you're delivering in that R&D function that you hoped you would capture when you were building those solutions. Pricing becomes a tool that you use to assign the right value and sell at the right value with the customers. We have been doing a better job at pricing. We have a much more sophisticated and robust process of pricing. But really, from our point of view, it's reflective of the value of the solutions that are being delivered.

So I would equate the growth in gross margin, approximately 300 basis points over the last 3 years. A lot of that has been on pricing, but it's not -- I don't look at it as we've raised the price. I look at it as the solutions are more valuable, and we're producing more valuable solutions. Also, as everybody knows, you can see that there's more value in the product because things cost more. So aluminum, steel, chips, labor, costs are going up. So we do have to adjust prices going forward, and we're improving our ability to do that.

Great. Let me stop here and see if there are any questions in the audience. If you have a question, please raise your hand.

Just wondering on the -- when you talk about your 8-quarter forecast, you said that they've been going up every time you've reported this year. Can you share what you're allowed to share on the shape of that forecast over time, right? Is it going up in a linear line? Is it tailing off? And to what extent is that just a reflection of lack of certainty from everybody involved in terms of their ability to project demand, say, 12, 18, 24 months out from now?

Okay. Good question. So on the 8 quarters, I wouldn't call it linear. And the reason is the way that, that comes into us, as you might imagine, is by factory. And so when factories open up, the size of the factory matters. Some -- one factory might be 30,000 wafer starts a month, another factory might be 50,000 wafer starts a month. There could be a mega factory that does even more than that. So as those factories become available to take tools and have tools installed, that translate into demand for us for those tools.

So I would say if you got right down to the smallest level of detail, it's very uneven in terms of its growth. It grows in spurts from that perspective. That said, our customers know that we don't produce in spurts. We produce relatively linearly from a tool perspective. So they try to schedule that demand with us as we grow. I would just say from a guidance perspective, we have said that we expect next year to be a strong growth year for the business, and it's really this AI investment cycle that we think is driving that. We don't see a change in that.

Generally speaking, we expect growth every year for semiconductors. We've looked in the past, we've talked about secular growth for semiconductors. We see capacity being added every year. We see the customers signaling that. When I talked about the fab tracker, the number of factory projects we're tracking globally, that's going up in each of the past few quarters, which gives you an indication that the amount of production capacity by wafer is expected to increase and customers are making those investments. So I guess the best I can say is it's not completely linear, and we don't see a place where it peaks, if you will, or changes from that perspective.

Brice, specifically about DRAM. There have actually been very few greenfield fabs that have come online over the last 5 years, and the industry is projecting 15, 16, 17 new fabs coming online out to 2030. Can you discuss the difference in the economics? And since they've been growing bits by primarily doing line enhancements and upgrades for the last 5 years. The next 5 years, we're going to be transitioning to greenfield. Can you talk about how the economics from your standpoint, specific to DRAM is different when you're going from line transition to greenfield fab space?

Sure. Thanks for the question. So on the DRAM side, I think a year ago, if I'm right, about 1.6 million DRAM wafers per month is a good number to think about. I think this year, the industry is adding 400,000 more wafer starts per month in capacity. And that's what we see going forward is that the next several years, you're going to get 300,000 to 400,000 additional wafer starts a month of DRAM capacity. And those are new factories, as you say.

The economics for Applied Materials and for any of the equipment companies is a lot different. I would say for an upgrade factory, just at a high level, about 25% of the equipment investment relative to what you see at a greenfield. So if the greenfield -- if the process equipment is more like $10 billion for 100,000 wafer starts of capacity for an upgrade, it's only 1/4 of that in that ZIP code. So this is a much larger investment cycle, which -- with many more wafer starts added per month in terms of production.

And that's a big difference you see in the DRAM cycle, so you can picture that, 1.6 million, 2 million, 2.4 million wafer starts per month. That's a completely different picture than what you see in NAND. NAND has been a smaller number of wafer starts and declining, even though the bit growth rate is the same, declining on a yearly basis. And that's because it's mostly an upgrade cycle for NAND and where you get those dynamics. It's less investment, less greenfield in that area. Thank you.

Brice, let's talk about your services business. It's growing 20% this year, a very strong year, above your mid-teens long-term model. Can you help us understand what's driving such strong services growth and why are you still expecting it to kind of come down to the low teens model?

Sure. So I guess the way I look at it, Atif, is the beginning of the year or the first half of this year, we talked about our services business growing in low double digits. So let's say, 11% or 12%. And we've raised that to mid-teens for a longer-term growth this year because the installed base is growing faster than we expected. It goes back to the semi business and what we were talking about with the semi business. Since the semi business has sped up, we're shipping more tools, the installed base for service is going up, and we have more opportunity there. So we raised our long-term forecast from low double digits to mid-teens.

This year, we're over 20%. And what's happening this year is that the utilization across the entire network has increased across every single end device. So leading-edge logic and DRAM, almost 100% utilization across the board. ICAPS is in a much more healthy spot and NAND has very high utilization. And so what that means is customers are buying more spares and more components to keep their systems working at full output. And so that's given us an unusual bump. Let's say we're at 100% utilization this year. We can't go to 110% next year. So you kind of next year go back to your more longer-term growth rate.

And just for the investors, the way I would think about the services business is there are 2 components to growth. One is that installed base growing every single year. 5%, 6%, 7% every single year. And then the second is the average revenue per tool rises each year as we introduce new products on those tools. Most of the new products or a large number of the new products are actually AI-based products. This is one of the benefits that we're getting from AI is we're able to provide tuning solutions to our customers that with the sensors that we have and with the data that we collect on the tools, we can quickly get those tools up to the highest yield and highest output.

Those are new services we offer to customers, and it helps with the revenue per tool even as that installed base grows. So that's the way to think about the service business. And obviously, mid-teens outlook is very positive for us.

Great. And then, Brice, when you talk about the value-based pricing approach, is that both on new products and existing products or mostly on the new products?

The way I would think about it is on every product, new or old from a -- maybe I shouldn't say we're old, but from an existing product perspective, your costs have gone up. So cost of labor, materials, components, aluminum, steel, gold, all the input costs have gone up, and that tells you that there's more value in that tool than there used to be. And so we have to make price adjustments on everything.

And I would just go back to kind of what we described a few minutes ago. We have a process to examine the value that the tool is providing. Each tool has a different level of solution capability and enabling capability for the customer. We look at that and we set the appropriate price. But I wouldn't distinguish between new products and existing products, we have to make an adjustment each period.

All right. And then on your process diagnostics and control, your outlook to grow more than 50% is actually far above WFE and also above some of your peer commentary in this market. Can you just talk about what you're seeing in that market?

Very exciting in process control. So if you can imagine, these are 3D architectures when you think about a Gate-All-Around transistor, many levels of semiconductor processing to build one of those processors and one of those transistors -- and you have -- you need the ability to look inside that construction. There's buried defects inside that construction, which optical sensors or optical microscopes don't allow you to see the e-beam microscopes where we have the -- or e-beam inspection where we have the top market share.

That is the tool that allows you to shoot electrons in there and see the shapes and see what's being constructed and find those buried defects and fix those buried defects. Our CEO says, you can't fix what you can't see. Well, that's one of our capabilities that we provide to customers. So as the architectures -- the actual device architectures themselves get more complex, you need more technology like the e-beam tool to look inside as you're building them, find the structural defects and be able to fix those in your process.

So the demand for e-beam is going up significantly, and we expect that to continue as DRAM goes to 3D DRAM, 4F squared and eventually as the leading-edge node goes from Gate-All-Around to CFET, you're going to need a lot more levels of inspection. That's been the #1 driver for us. We're also making several other investments. I mentioned the X-ray. I mentioned the package-level e-beam products that we have. So we think the inspection business will be significantly important for Applied Materials and its growth story.

Great. That's a wrap. We're almost out of time. Brice, thank you for coming to the Citi conference.

Thanks, everybody. Good to see you. Thank you.

Welcome to the Applied Materials Third Quarter of Fiscal 2026 Earnings Call. [Operator Instructions]

I would now like to turn the call over to Mike Sullivan, Corporate Vice President of Investor Relations. Please go ahead.

Good afternoon, everyone, and thank you for joining today's call. With me are Gary Dickerson, our President and CEO; and Brice Hill, our Chief Financial Officer.

Before we begin, I'd like to remind you that today's call includes forward-looking statements, which are subject to risks and uncertainties that could cause our actual results to differ. Information concerning these risks and uncertainties is discussed in our most recent Form 10-Q and other filings with the SEC.

Today's call also includes non-GAAP financial measures. Reconciliations to GAAP measures can be found in today's earnings press release and in our quarterly earnings materials, which are available on our website at ir.appliedmaterials.com.

In addition, any comments regarding calendar 2026 refer to Q2 of this fiscal year through Q1 of fiscal 2027, which will be a 14-week quarter.

Next, I'd like to remind you about our 2 special events during SemiConwest. On Monday afternoon, October 12, we will host an unveiling of the new EPIC Center in Silicon Valley, California. And on Tuesday morning, October 13, we hope you'll join Gary, Brice and our business unit leaders for our investor breakfast presentation at the Yerba Buena Center in San Francisco. You can join us in person or on a live webcast.

And with that introduction, I'd now like to turn the call over to Gary Dickerson.

Thank you, Mike. In our third fiscal quarter of 2026, Applied Materials delivered another set of record-breaking results, including the highest quarter-on-quarter revenue growth in the company's history. The rapid global build-out of AI infrastructure combined with Applied's leadership positions in the most enabling and highest value technologies for AI computing provides the company with an exceptionally strong foundation for multiyear revenue and profit growth. As 2026 has progressed, customers have found new ways to address clean room space constraints and significantly increased their demand for tool deliveries. In the past three months, we have again made upward revisions to our revenue growth forecast for the year, and we are confident we will grow faster than the overall market.

As AI computing drives unprecedented demand for semiconductors, there is a large gap between demand and supply for advanced chips. To ensure our supply chain and field teams can support their ramp, our largest customers are giving us longer-term commitments and rolling 8-quarter forecasts. This increased demand visibility gives us high confidence that 2027 will be another strong growth year for Applied Materials. As customers move quickly to bring new fab capacity online, while simultaneously optimizing yield and output of their existing production facilities, we also see strong incremental demand for our advanced service solutions.

In my prepared remarks, I will share my views on how AI is resizing and reshaping the semiconductor industry and its ability to realize the value of advanced technology. I will describe the increasing value applied to delivering to our customers by accelerating their technology road maps, optimizing existing production capacity and helping ramp new fabs faster. And I'll provide a brief update on our EPIC strategy as we prepare to start operations in our new EPIC Center in Silicon Valley.

As I have said before, I strongly believe that AI is the biggest and most consequential technology inflection of our lifetimes. While we are still in the early innings of deployment, AI is reshaping the global economy and becoming fundamental to the relative competitiveness of companies. What I am seeing at Applied Materials is a great case study in AI's real-world impact. Our investments in AI are on track to deliver compelling returns by accelerating our revenue growth and operating profit margins. In R&D and services, we are using AI to create highly differentiated products, significantly speed up product development time lines and create valuable new service solutions for customers. In operations, supply chain and our corporate functions, AI is helping us ramp faster, drive meaningful improvements in productivity and scale our revenue significantly faster than our head count.

Beyond Applied, we see similar trends playing out across a wide variety of industries. As the performance and cost of AI computing improves, many new applications will become technically viable and economically attractive. These expansive opportunities for value creation are fueling an intense global competition for AI leadership, which can be described as two concurrent races. The first race is for technology leadership. AI data center returns are determined by the number of tokens generated per second and the total cost of ownership, which is dominated by energy consumption. Improvements in token per second per watt are primarily driven by innovations in semiconductor devices and systems. This can be seen in the value of the semiconductor and semiconductor equipment industries are generating from their most advanced technologies.

The second raise is for capacity as demand for advanced semiconductors to support AI infrastructure scaling far exceeds supply. As a result, chipmakers are intensely focused on increasing output and yields of their existing factories while rapidly building new ones. For Applied, the technology race and the capacity race are fueling new opportunities to create and capture value. In the race for technology leadership, leading edge foundry logic, DRAM and advanced packaging have the greatest impact on AI computing performance, power efficiency and cost. Together, we expect these areas to represent around 80% of wafer fab equipment growth in 2026 and 2027. These three areas are also where Applied has strong leadership positions, where we identified key AI inflections early, and where we've shifted our investments to build an innovative pipeline of next-generation solutions.

In the past quarter alone, we have announced six new products, including our Centura Prime Epitaxy system, designed specifically for high-performance DRAM, producer Avila that enables higher performance and higher layer count high-bandwidth memory, Nokota VMAX, our next-generation plating system and Opta Quad CMP for advanced packaging and two new e-beam systems also for advanced packaging that expand on our e-beam leadership in the front end.

Advanced packaging is one of the most important areas for AI compute innovation, and we see very strong multiyear growth for Applied. Applied as the overall leader in this market with strong positions in high bandwidth memory and 3D chiplet stacking, and we now expect our overall packaging revenues to grow more than 70% in calendar 2026. We are also well positioned for future packaging inflections as the industry moves to new architectures and larger-sized panel formats. We have built a broad portfolio of next-generation technologies for panel, including digital lithography, deposition, etch and e-beam review.

In the global race to add semiconductor manufacturing capacity, our customers' ability to increase yield and output in their existing production fabs is incredibly valuable. This creates expanded opportunities for Applied to deliver new innovations in three key areas: Services, process diagnostics and control, and new products that increase wafer output per area of fab space. Our advanced service solutions enable customers to better optimize performance of their high-volume manufacturing operations. We already have more than 37,000 chambers in the field connected to our proprietary AIX software capabilities and use AI-powered monitoring, diagnostics and predictive analytics. Our advanced services are delivering yield improvements for customers and helping us drive higher growth rates in applied global services. We now expect AGS to grow more than 20% in calendar 2026 and to deliver a sustainable long-term annual growth rate in the mid-teens.

Our metrology and inspection product portfolio is also enabling customers to accelerate fab yields and output improvements. The most advanced logic and DRAM devices require more e-beam steps that can provide sub-nanometer resolution for high aspect ratio structures. Applied has unique e-beam technology and is the leader in this growing market. In parallel, we are introducing new optical inspection products, which enable us to increase application share in these markets as well. As a result, we expect to grow our process diagnostics and control business more than 50% in calendar 2026, and we have a strong pipeline of new products that will fuel growth in 2027 and beyond.

Finally, we are developing a new portfolio of output innovation products that increase the wafers that can be processed per square foot of clean room space. One example is our new Epitaxy system for DRAM that not only increases device performance, but also uses 20% less clean room space than our earlier products. We have multiple output innovation products in qualification at customer sites that will provide significant increases in output per unit area. With incredibly strong customer pull for next-generation technology and unprecedented demand for semiconductor manufacturing capacity, the value of time to market has never been greater.

Our EPIC strategy is designed to increase innovation and commercialization velocity by creating earlier and deeper engagements with our customers and partners and co-locating key innovators. For chip makers, EPIC provides much earlier access to Applied's new product innovations that are at the foundation of future AI compute architectures. The output from Epic will be more mature technology that can deliver high yields faster in volume manufacturing. For Applied, EPIC co-innovation programs will enable us to be designed in to new chip and packaging architecture inflections, increased R&D productivity and value sharing and provide better multi-node visibility to guide our investments and resource allocation.

Since our last earnings call, we announced that Broadcom will join EPIC as an innovation partner to accelerate the development of advanced chip packaging technologies for next-generation AI systems. We also signed EPIC partnership agreements with screen and UC Berkeley. This brings our total number of announced EPIC engagements to 11, spanning system companies, leading chip makers top research universities and innovation partners. The center pet of our EPIC platform is our brand-new EPIC Center in Silicon Valley. We will move the first R&D tool into the clean room next week and we are on track to start operations in the coming months.

Before I hand over to Bryce, let me briefly summarize. Demand for advanced semiconductors and semiconductor equipment continues to strengthen, and as customers find new ways to address clean room space constraints, we see higher demand for 2026 tool deliveries. With support from our supply chain, we have again increased our expectations for 2026 revenue, and we are confident we will grow faster than the overall market this calendar year.

In the race for AI technology leadership, leading-edge family logic, DRAM and advanced packaging have the greatest impact on AI computing performance, power efficiency and cost. These are areas where Applied has strong leadership positions and an innovative pipeline of next-generation solutions, supporting strong revenue and margin growth in 2027 and beyond. And we're working closely with our customers to optimize yield, outputs and fab ramp times with valuable new innovations in services, process diagnostics and control and output innovation products.

Brice, over to you.

Thanks, Gary. I'm pleased to share that Applied delivered another quarter of double-digit sequential and year-over-year growth in revenue, operating profit and non-GAAP earnings per share. Fiscal Q3 also marks our 13th consecutive quarter of year-over-year gross margin expansion, which demonstrates how we are benefiting from the tremendous value our products and services bring to our customers and the entire AI ecosystem.

Our fiscal Q4 guidance demonstrates continued strong year-over-year momentum. And in the second half of the calendar year, we expect particularly strong growth in DRAM as well as leading-edge foundry logic and the advanced packaging for both. On today's call, I'll update you on the [indiscernible] environment, discuss how we're scaling our operations for continued growth, demonstrate how value creation is expanding our gross margins, summarize our Q3 results and provide our Q4 guidance. Over the past quarter, the demand outlook has strengthened across all the leading indicators we track. Cloud service providers continue to increase their investments in AI infrastructure. Importantly, many of these companies are already generating positive returns on their investments. And so are their enterprise customers, including Applied.

As Gary described, we are accelerating the pace of new product development, increasing revenue and generating new efficiencies in our support functions. In fact, G&A as a percentage of operating expenses has declined to the lowest level in our history.

Turning to our direct customers. Most leading-edge logic and DRAM fabs are running at full capacity. Utilization levels are rising across the board, including an ICAP where we see strong demand in AI-related markets like power and optical chips. As a result, our customers have announced more than 10 new fab projects just this quarter. Customers continue to give us longer visibility than we've ever had with some conversations now extending to 2030. These communications are valuable to our company and our own supply chain partners who are scaling with us to support our customers' growth forecast. During the quarter, we officially opened our newest manufacturing center in Singapore and combined with other expansions worldwide have nearly doubled our manufacturing space over the past several years. Based on the longer-term demand signals from our customers, we are now taking this further hiring and training new manufacturing and customer support teams so that we have the capacity to double our quarterly system output from current levels by 2028.

In fact, we add more than 1,500 people this quarter in worldwide manufacturing and AGS customer support. We are also planning our next manufacturing capacity expansion, ensuring we have the option to support further increases in demand by 2030.

Next, I'll discuss value creation and sharing. Applied is delivering value to our customers in the AI ecosystem in more ways than ever before. We've increased R&D every year since Gary joined the company in 2012. And the investments have broadened from equipment innovations to materials engineering solutions that result in better chips. More recently, we've significantly increased R&D and advanced packaging innovations that enable better systems supplementing our R&D with two small acquisitions. And today, we're increasing investments in technologies that enable better fab economics as we accelerate ramps and boost output and yields. All of these technologies will be put to work at the EPIC center, where we will co-innovate with our customers and partners to accelerate the AI road map.

In short, we've broadened our focus from making better equipment to enabling better chips and systems for AI and better fab returns for our customers. These investments have made us a more valuable partner to our customers and enabled us to share in the value we create. Three years ago, we implemented a systematic approach to value-based pricing. And today, you can see the benefits reflected in our strong revenue growth and gross margins which have increased to over 50% for the company and over 55% in semiconductor systems. We have higher pricing and margins in both new and existing products. At the same time, we remain focused on cost improvements and use them aggressively to help offset higher input costs. As we look at the many opportunities we have to further increase the value of chips, systems and fabs, we are confident we will continue to expand gross margins.

Next, I'll summarize our Q3 results. We generated record revenue of $9.1 billion, up 15% sequentially and 25% year-over-year. Non-GAAP gross margin increased to 50.4%, up 40 basis points sequentially and 150 basis points year-over-year. Non-GAAP operating margin expanded to a record 34%, up 190 basis points sequentially and 330 basis points year-over-year. And we delivered record non-GAAP earnings per share of $3.50, which is up 22% sequentially and 41% year-over-year. Last quarter, I discussed our focus on increasing operating leverage. We grew revenue much faster than spending in Q3 on both a sequential and a year-over-year basis and drove OpEx as a percentage of revenue to the lowest level in nearly 4 years.

Turning to the segments. Semiconductor Systems delivered record revenue of $7 billion, which is up 18% sequentially and 27% year-over-year. The revenue mix was similar to last quarter as capacity additions and gate all around and FinFET, drove record foundry/logic revenue. DRAM revenue, which includes HBM packaging, grew by 52% year-over-year to record levels. As we look ahead to the second half of the calendar year, we expect a very significant increase in DRAM revenues as our customers begin to expand clean room capacity. Looking to our individual materials engineering business units, we had record revenues and deposition in Q3, including in PVD, CVD and epitaxy, which is one of our fastest-growing businesses this year. In materials modification, we had record sales in thermals and treatments. In materials removal, we had record revenue in both etch and CMP. We also had records in process diagnostics and control, which is growing faster than our overall systems business this year. Segment non-GAAP gross margin increased 190 basis points year-over-year to 55.4%. Non-GAAP operating profit increased 45% year-over-year to a record $2.7 billion.

Applied Global Services delivered record revenue of $1.8 billion, which was up 22% year-over-year reflecting both subscription services growth and high transactional parts demand. Ramp readiness is a major priority, and the team added more than 1,000 customer support engineers. AGS is using AI and warehouse automation to grow as efficiently as possible, which is reflected in strong gross margin of 35.6%, up 180 basis points year-over-year and operating margin of 30.1%, up 280 basis points year-over-year.

From a regional perspective, China represented 26% of our semiconductor systems plus AGS revenue. We now expect our China revenue to increase this calendar year, led by investments in 28-nanometer foundry logic, where Applied has strong technology differentiation and share. Other revenue of $294 million was in line with our expectations. We generated record operating cash flow of over $3 billion. Capital expenditures were $707 million, resulting in free cash flow of $2.3 billion. We distributed $860 million to shareholders, including $420 million in dividends and $440 million in stock repurchases. We have $12.8 billion remaining in our share buyback authorization and continue to expect to distribute 80% to 100% of free cash flow to shareholders.

Now I'll share our guidance for Q4. We expect company revenue of $10.25 billion, plus or minus $500 million, which is up 51% year-over-year. We expect non-GAAP EPS of $4.02 plus or minus $0.20, which is up 85% year-over-year. Within this outlook, we expect Semiconductor Systems revenue of around $7.9 billion, up 62% year-over-year, AGS revenue of about $1.84 billion, up 22% year-over-year and other revenue of around $510 million, composed primarily of display revenue. I've said previously that our display business includes new products that could help us drive higher quarterly revenue in future periods. For modeling purposes, we now expect other revenue to be approximately $400 million per quarter on average through 2027.

We expect non-GAAP gross margin to be approximately 50.4% in Q4, up 230 basis points year-over-year. And we expect non-GAAP operating expenses of around $1.58 billion. As a reminder, Q1 of fiscal 2027 will be a 14-week quarter which will result in a higher than average step-up in our Q1 operating expenses. Finally, we are modeling a non-GAAP tax rate of around 11% and and a tax rate of around 13% in 2027 as we absorb the effect of the global minimum tax.

In summary, the rapid adoption of AI that we've been investing for is driving strong growth and record revenue and profitability for Applied Materials. We are enabling better chips, systems and fab returns and systematically sharing in the value we create. We see continued record performance in the second half of the calendar year with a sizable increase in DRAM and leading-edge foundry logic revenue. Based on the unprecedented visibility we're receiving from our customers, we expect another strong record year in 2027 and are making substantial investments to be able to ramp to higher levels beyond next year.

Now Mike, let's begin the Q&A session.

Thanks, Brice. [Operator Instructions] Operator, let's please begin.

[Operator Instructions] And our first question for today comes from the line of C.J. Muse from Cantor Fitzgerald.

Quarter ago, you quantified semi-systems growth of 30-plus percent. Curious if there's kind of a framework for thinking about what the growth outlook looks like now, given your positive commentary. And is there any sort of framework that we should be thinking about into calendar '27?

Hi, C.J., it's Brice. Thanks for the question. So our key comments there and the way that we're seeing the business is that demand strengthened again during the quarter. We see new projects being added by our customers on factory side. We see CapEx forecast going up by our customers, and we see strong CapEx from the cloud service providers all and out. So the mid -- the greater than 30% that we highlighted last quarter, we're saying now that it's greater than that at this point. We don't want to guide our out Q1 at this point. So that's as much information that we're providing. But when we look into 2027, we expect this whole demand function led by AI to continue. So we're saying '27 at this point, we expect another strong year.

Yes, C.J., this is Gary. I would add just that the fastest-growing parts of the market are the leading-edge foundry logic, DRAM and advanced packaging. We said that's around 80% of the growth in wafer fab equipment spending this year, and then we'll see a similar profile in '27. Those are the fastest-growing parts of the market. Those are areas where we have clear leadership and really well positioned going forward. So as you said, we increased greater than 20% is what we said year-over-year in February, then greater than 30% in May. And now we think stronger than that, going forward. And so demand in all the conversations with the customers is very strong. And what we're hearing from customers with these 8-quarter rolling forecast is very strong multiyear demand. I really want to thank our supply chain teams and operations teams. They're doing a great job in responding. But again, customers are being very creative in how they're able to expand space and take tools earlier. Our teams are reacting very quickly, and we're in a strong position to outperform this year.

So strong environment, C.J., and we also highlighted that we expect to gain share during the year.

Very helpful. And I guess as a follow-up on gross margin, you talked about value-based pricing. And I think we've heard from most companies around anything expedited service new tools. But you talked about like-for-like pricing pushing higher. So could you speak maybe directly to that, and how we should be thinking about the implications to your overall gross margins as we proceed into fiscal '27 and beyond?

Sure. So over the last three years, approximately 300 basis points increase in our company level gross margins. One driver for that was our value-based pricing that we do for every single tool. So the issue C.J. was when we came through COVID, supply chain crisis, cost of inputs went up, we found ourselves needing to reprice every tool. So we put that value-based process in place. And that's what we do is examine the value of every single tool and put a new price on every single tool. And we think that's required in an environment where the input cost change constantly. And then looking forward, we expect to be able to continue to improve our gross margins. We're already over 55% for our -- at the semi systems level and value-based pricing will continue to be a part of that.

Yes. C.J., I would add that I think our ability to create value for customers has never been stronger. This race to bring new AI architectures to market is what every one of our single customers is focused on. Applied has the most enabling technologies, as I said, in the fastest-growing parts of the market. So we have a tremendous opportunity for creating value there. And I also talked about yield and output innovation. So all of our customers, they are racing to be first to market with these new architectures and then also ramping as fast as possible optimizing yield and output. So that puts us in a position where our products are extremely valuable. Our services are also more valuable. That's driving the greater than 20% service growth that we're seeing. We talked also about greater than 50% growth in our PDC business that also is related to yield optimization. And then our pipeline of new products is also very strong, and all of those have higher margins, and that will give us a tailwind going forward.

And our next question comes from the line of Vivek Aria from Bank of America Securities.

So Gary, many of your memory customers are saying that they have 3- to 5-year long-term agreements with good visibility into units and pricing? I know you mentioned you have 8 quarters of visibility. But as you look at your customers who are signing up for a greater level of alignment with their end customers. How is that kind of translating into your longer-term visibility beyond these 8 quarters?

Yes. I can start on that, [indiscernible], it's Bryce. So what we have from customers, of course, the large customers, we actually have visibility to the road map. So we have a perspective on probably 5 years of visibility for our largest customers. we asked them for the detail at a detailed level for the 8 quarters that we've been speaking about, so we can aggregate that and get it to our supply chain. And other things that have changed, we get longer lead time POs from our customers so that the details are agreed from a longer lead time perspective. And then we also have some charges that have moved into this environment like cancellation charges and expedite charges that help will -- that help navigate the environment. So there's been a number of changes, and I think visibility is significantly increased from prior periods.

Yes, Vivek, this is Gary. Certainly, everybody can see that there is a gap between supply and demand, DRAM, especially with AI as we're expanding from training to inference to a genetic AI to physical memory demand continues to go higher, and I'd say especially DRAM. DRAM this is going to be a very strong growth year for Applied, more second half weighted in terms of our DRAM growth, but very strong growth in DRAM. And I think, as you know, we've expanded our DRAM share significantly over the last several years, and we continue to 'see strong growth in our DRAM business going into 27. And as you said, as we're in all of these conversations with customers, they're talking about multiyear growth and significant demand that we're increasing our capacity to meet -- and then for us, relative to our position in DRAM, we are the #1 process equipment provider in DRAM. We're the leader in the CMOS periphery logic to upgrade transistors for higher performance and power including, I talked earlier on the call about very strong epi growth, HBM packaging. We're the leader in materials deposition for wiring and patterning, conductor etch e-beam technologies, and we're in deep partnerships with customers also for future DRAM architectures, innovations in 6 F squared, 4 S squared, and we're very well positioned for 3D DRAM in the future. So I have high confidence that we're going to continue to drive significant growth and gain share in this segment.

Got it. Form a -- follow up, one more on gross margins. If I go back from calendar '21 to '25, your gross margins and those of your U.S. peer were about the same. In fact, Applied was slightly ahead. But in the near term, they are about 150 basis points higher. So I understand mix for every company is different. But I'm curious how come the gross margins were so aligned in those 5 years, yet they are kind of lagging a little bit. So just what's the prospect rise for expanding gross margins to more of this industry level going forward?

Yes, Vivek, I don't know if I have a great reason to think about comparisons. But from our perspective, we've made tremendous progress from a gross margin advancement. We talked about our value pricing and really it's the portfolio. The portfolio continues to strengthen as we target the R&D and the collaborations with our customers at the most valuable inflection solutions that need to be developed. And so the pricing really is just demonstrating the value of that portfolio. And yes, we expect to be able to continue that. Of course, we have some other elements in our portfolio like the display business that we've talked about when that grows faster, that has a different effect on the corporate gross margin. So the portfolios are different between the two companies, but we expect to be able to continue to improve our gross margin and grow the value over time.

Yes. again, I would also add, we have been driving margin growth, pretty significant margin growth. We talked about 13 consecutive quarters of year-over-year growth. 90 basis points in our semi business in the last year. And I have high confidence we're going to continue to drive margins higher, continuing the growth that we've seen over the last few years.

And our next question comes from the Stacy Rasgon from Bernstein Research.

I have one more on gross margins as well. And look, I feel a little bad harping on it because they're actually really good. They're well over 50%, and they are coming in higher than you had expected. But I'm just wondering, you are guiding them flat at these current levels, at least in the near term on a pretty sizable revenue increase in just given all the commentary around portfolio and pricing and everything, I was just a little surprised. Why is it? Is that just a function of the display business growing sequentially just mix between the businesses? Or is there something else going on in the near term?

Hi, Stacy, Brice. Yes, thanks for pointing out the growth in display is certainly a factor in the recipe. But really, it's just a ramp headwinds as with the growth, we're ramping a lot of customer service engineers. We have a lot of resources that we're adding in the semi business. And so yes, we get the benefit of strong segment mix with semi growing so strongly and we get the benefit of more volume. But we do have some ramp costs that are in the forecast. So we feel good about in our guided quarter, we feel good about having a flattish margin at the company level. And then like we said, as you look longer term, we'll expect to be able to continue to grow the margin.

Got it. So I guess to follow up on that then. So if there's ramp costs now, I guess how long do those last? Are you still ramping up more engineers and other costs like into subsequent quarters? Or are you all in this quarter? And then as we think about the pace of that expansion, like what does it look like? I think you talked about last quarter thinking about something like 10 bps or something sequentially kind of going forward. Is that still the trajectory or given some of the other drivers, do you think it can come in better than that?

No, I think the continued pace of improvement, we called that slow in the past. I think slow improvement is the right way to think about it. We do expect to improve it over the longer horizon. So I think we will continue to add employees over the next few quarters, but I think that headwind will recede as the revenues continue to grow.

our next question comes from the line of Timothy Arcuri from UBS.

Brice, I want to go back to the systems guidance for the year. So you said up more than 30 last call, things have gotten better since then. But even if you go 40%, that implies a pretty big decel, into December. So you go from 18% in July to 12% in October to then 6% in January, and that gets you to like 40%. And Lam was talking about WFE being up like 38%. So if you're going to outgrow you have to be at least 40%, if you believe their numbers. So would you commit to be to growing systems 40% or more?

Hi, Tim, thanks for the question. We are committing that we expect to outgrow. We think that's already happened so far this year, and we expect to have that happen. Wherever we land from a growth perspective as we get through the year. And I think you've got the dynamics right. We first guided greater than 20%. We raised that after customers added clean room projects. We raised it to greater than 30%, and we're saying it's even higher now. And so yes, we'll expect -- we're not giving that number because we're not guiding that out quarter, but those are the right dynamics. I guess in the last I guess the last thing I'll add there is we do expect sequential growth in our Q1, the calendar Q4, but not guiding that at this point.

Okay, Brice. And then maybe asked a different way. So the comment about manufacturing capacity doubling. So should I just take that kind of face value and so you're shipping roughly your revenue in roughly $7 billion in the July quarter. So at a value does that mean that sometime during calendar '28, you'd be revenue at $14 billion, or is it more newer than that?

It is more nuanced. I think you should take it explicitly. It's capacity. So it's not a revenue forecast for 2028. What we have to do with long lead investments like clean room is make sure we have the clean room in place with a profitable business or any demand forecast demand reality in that environment. And so we're just communicating because partially we're communicating to our suppliers also we're putting the capacity in place to be able to support a wide range of output requirements in 2028. And as far as that goes the years after also. So no, it's not a revenue forecast, but yes, it's giving you an indication of what we will be prepared for.

And our next question comes from the line of Krish Sankar from TD Cowen.

I just want to follow up on Tim's question. Gary, or Brice, it seems like your customer conversation seems to have shifted from annual price discussions to basically delivering and meeting requirements basically time to market with a two-year visibility. I'm just wondering, does you give you some freedom on pricing? Or does it add more burden on expenses on prepping up your capacity underspend the supply chains at the upcoming ramp. And if a visibility of 2 years, why not give a Jan quarter [indiscernible] the outlook? And then I have a follow-up.

Hi, Krish, yes. The customers definitely -- especially the large customers, we've got very strong visibility. And you're right, they are becoming more and more interested in scheduled delivery and hitting the schedules, which is why they're collaborating, cooperating on specificity of the orders and the 8-quarter visibility that we're also sharing with our suppliers. So I think that dynamic has improved the situation for us in the planning environment dramatically from the prior year. So I think all those things are true. I guess that's our perspective at this point.

Got you. Got you. And then maybe, Brice, think in your prepared comments, you kind of spoke about certain customers giving you visibility into 2030. I'm just curious, are those connotations about technology, or is that still about capacity to scale up and meet that demand?

Well, yes, it's definitely about technology. So especially with the large customers for mature customers. We know the fab projects that are on the road map. We know the technology that's planned and even if it's a new technology, we have a perspective of our positions in those technology. So we do have the ability to plan in a detailed manner for 5 years. And then what happens at the 8 quarter out is we get very specific about the node and the tool types that need to be built so that we can pass that information along to the suppliers. And we do have...

Yes, Krish, we are -- I mean many of those conversations with the CEOs of our largest customers. And for sure, they're giving us visibly detailed visibility for the 8 quarters, but even beyond that, I think they have communicated they see strong multiyear demand in their business. And so they're wanting us to be ready to support that demand. And as you know, it takes time for us to get the supply chain ready to support those levels. So we are getting the visibility relative to capacity needs beyond the 8 quarters. On the technology, I would say that those discussions go out maybe 10 years in the future because Applied is the most enabling for those key architecture inflections. We have the most broad, the most connected, the most unique portfolio to enable the new transistors, the wiring, the DRAM architectures, new packaging architectures and it takes time to bring those innovations to market. So Applied is unique in the portfolio we have. We have the most enabling technologies. If you looked at the top or technologies you need to enable these new architectures. Applied has, by far, the majority of those technologies. And so very deep co-innovation relationships with customers where we're cocreating those architectures. So that technology visibility goes beyond 5 years in these deep co-innovation relationships with the customers.

And our next question comes from the line of Harlan Sur from JPMorgan.

The team previously guided your global ICAP business to be flat to slightly this year. Outside of China, I mean, we are seeing a strong cyclical recovery in automotive, industrial, for your analog power microcontroller customers right? They are also articulating an environment of tight supply as well. Utilizations are rising meaningfully, Brice, as you articulated. But are these global customers also starting to pick up their spending? And do you see your total ICAP's business growing this year?

Harlan, thanks for the question. So I'll start here. Yes, in pointing this out, we do see a change in the ICAP dynamic. So you called out the increasing utilization across those customers. That's a positive. Our view of China is that it will grow this year, and it will grow next year. That's a big part of our ICAP portfolio. And we do think that ICAPs overall will grow this year and will grow next year. So the digestion we're hopeful that the digestion that we've talked about on the equipment side in the past is expiring, and we can return to growth. And then specifically, for our non-China customers as we look into next year, we think there'll be positive growth for those customers. We see bright spots in power and photonics in other areas. So it's looking much more positive than it has the last couple of years.

I appreciate the color there. And another quarter of strong revenue growth and growth in operating margin profitability in AGS, right? 30% operating margins, I think that's the highest level, I think, in two years. Plus your incremental gross margins, I think for the last couple of quarters have been low above 40%. So strong and also have been delivering strong incremental operating margins as well. I know the team has historically thought that they could drive AGS operating margins longer term into the low 30% range. But on your strong incremental margin profile, could we see gross margins approaching the 40% range and operating margins in the mid-30s as AGS revenues continue to scale higher advanced services becomes a bigger part of the mix kind of over the long term?

Yes. Thanks, Harlan. So I think not different from the semi business. We do expect that we have the opportunity to improve gross margins over time in the services business. And what's happening there is that solutions like the information solutions that come from AI are allowing us to develop new products and be more efficient in the services that we're providing. That combined with the growing installed base gives us good growth there. And then this year, we had the benefit of significant increase in utilization, which grows the spares business more quickly than in prior years. So that helps us from a gross margin perspective. That -- from that point, you can only grow to 100% utilization once. So that sort of slows down. But we do expect we'll be able to improve gross margins over time in the services business.

Yes, Harlan, I would add that I would add that for customers right now, optimizing output and yield is incredibly important in a supply-constrained environment, and that's going to go on for some period of time. So the value of services that optimize yield is incredibly valuable. And the good news is we have a lot of new innovations there. We talked about over 37,000 chambers connected to our AX servers. We have AI-enabled applications for preventative maintenance or chamber matching. Those are incredibly valuable services that will drive our top line growth, our service contract growth faster and also enable us to capture value more quickly. So actually, I am more positive on growth in the AGS business, both top line and bottom line than I've ever been.

And our next question comes from the line of Blayne Curtis from Jefferies.

I want to ask on NAND. It doubled in the quarter, obviously, off a small base. Just curious what you're seeing in that market. And then you did mention it in October. I'm assuming maybe that's a small dollar that's why you didn't call it out, but is it continuing to grow?

Hi, Blayne. Yes, good growth in NAND this year from a percentage perspective, small base, like you say, but we think this year is a strong year of growth for NAND. As we look forward, into our next year, we think the dynamic is that really that AI dynamic that we've called out before and Gary mentioned earlier, leading-edge logic and DRAM and advanced packaging will be the fast growers. ICAP should return to growth for us, NAND to grow but will be the slower grower in the out year.

And then, Brice, maybe on CapEx. You talked about the expansion of spending for 2030. I don't know if that's going to hit now. I'm just kind of curious if you can comment on what you expect OpEx to be in October? And then just any perspective next year? I think EPIC rolls off. So I think the prior was that it would go down. But now with the strength in the business, I'm just curious how you're thinking about that spend.

Yes, that's a great question because you're right, on the strength of the business, of course, we have more additions that we want to do. This is a CapEx comment. We want to -- we will be putting equipment inside EPIC as well as other investments. So what I would say about CapEx, it will still be a year that's higher than normal, but it will decline as a percentage of revenue as we go into 2017. So that's our perspective at this point.

And our next question comes from the line of Jim Schneider from Goldman Sachs.

I was wondering if you can maybe comment, given the strength you're seeing across your focus areas. As you look into fiscal '27 or calendar '27, when you care to rank order where you see the incremental strength between foundry/logic, DRAM spending in advanced packaging?

Yes. Jim, this is Brice, I'll start. We actually don't distinguish between them. We think the system-level pull that AI provides is similar across those different end markets. So I wouldn't call the difference enough to distinguish between them. So it will be strong for leading logic. It will be strong for DRAM. It will be strong for advanced packaging. And what's new for us now is that we also think that ICAP will grow next year. So that's another difference from our previous 90 days.

Yes, Jim, thanks for the question. Yes, I think we talked earlier about 80% of WFE growth in those three segments in '26. And we see a similar profile, maybe even a better profile. But right now, we'd say similar profile in '27. Those three segments, as Brice talked about, we see those as the fastest-growing segments in '26 and '27 and frankly, over the next several years.

That's helpful. And then just given the first half versus second half dynamic we're seeing in calendar '26. Is there any reason why in calendar '27, you wouldn't see an accelerating growth rate for overall revenue. Is there anything you see on the horizon that would give you pause?

I think a lot of people asked Jim about what is governing growth. I think as you move into the longer term, what governs growth is clean room from our perspective. So customers continue to add clean room projects. That's why we raised our forecast this year. Some of those will add incremental clean room space next year. And of course, it gets larger as the out years come into into focus because those projects usually take a number of years. But anyway, I think it will be the availability of clean room at a high level that will determine what we can all ship next year.

And our next question comes from the line of Mehdi Hosseini from Susquehanna International.

Yes. I just have a couple of follow-up for Gary. Forget about the near-term trend, but I want to better understand how you're thinking about your targeted revenue, gross and operating margin for a scenario where WFE will be $150 billion to $175 billion. Any color would be great.

Hi, Mehdi, thanks for the question. Yes, I think when we have the October investor event, we'll give more color relative to growth rates. What I would say is that if I look at the setup relative to compute demand going forward. We see this as a strong multiyear growth driver and the fastest-growing segments market, what we talked about earlier, the most valuable parts of the market leading-edge foundry logic, DRAM, advanced packaging. We're #1, and we're positioned to gain share going forward in those segments. So I think top line growth, we have a number of really great drivers that will enable us to outperform in '26, and we look at very strong growth going forward. And I also believe that really across the board, the value that we're delivering is also increasing. We are -- we do have the most unique connected portfolio that's creating tremendous value for our customers, for new chip and packaging architectures, that puts us in a good position to continue to drive our margins higher going forward. And I mentioned earlier about the innovations and yield and output. And as you can imagine, Mehdi, every single customer, they're focused on getting as many chips out per square meter as they possibly can. So that increases the value of our service business, and we're bringing innovation that really directly address those areas of focus for our customers. So I think relative to the top line growth, bottom line growth, I'm very optimistic. But we'll give more color, Mehdi, when we have our October investor meeting.

I'll add -- I'll just do one add, Mehdi. You're really describing this year and scenarios that people are talking about for this year. So I think we're giving you the ingredients for that with a greater than 20% services business for the calendar year, something higher than greater than 30% for the semi business. We talked about our gross margin outlook. We give you the display item. And then whatever that WFE number is, we've said we expect to gain share. So I think you have the ingredients needed to have a perspective on that.

Got it. Just a quick follow-up. I think your NAND has been relatively the smallest part of your semi and I think it's a reflection of more of an upgrade going on within the NAND than wafer capacity add. When do you think the industry would actually start adding some with the capacity to offset some of these losses due to increased migration to a [ 300-layer ] account.

Hi, Mehdi. Yes, so for the dynamic here, I think we've described in the past is wafer starts continue to decline in NAND. So the projects you see are for upgrades, as you described, to get more layers. We expect the environment to continue the same dynamic for the next few years. The one place that will be different is new projects in China, I believe. But for the larger customers, it's mostly increased space to afford those upgrades in layers that you described.

Yes. Thanks, Mehdi. And operator, we have time for one more question, please.

Then our final question for today comes from the line of Srini Pajjuri from Royal Bank of Canada.

Gary, a couple of questions on the technology front. You talked about panel-level packaging. And just curious as to when do you see panel-level packaging becoming mainstream. And compared to your current position in advanced packaging, what sort of opportunities do you see in that market? And then also, if you can touch on hybrid bonding. It seems like that's finally happening and what sort of opportunities you're seeing in the market?

Yes. Thanks for the question. So as I talked about earlier on the call, packaging is one of the most important areas in the industry relative to improving AI compute performance and power, multichip connectivity value the data, huge focus for all of our existing customers and new customers that are innovating with new architecture. So over 70% growth this year. We have, by far and away, the strongest portfolio of technologies. We've also talked about acquisitions that we've done here recently that add to our strength in packaging. So I have very high confidence that we will continue to drive significant growth over many years in packaging. Relative to new substrates, all of our customers are really focused on connecting as many logic and memory chips together at the highest performance and power as they can. And so we're in deep co-innovation relationships with companies, and there is a race for all of these companies to drive these new architectures to market because there's so much value in performance and power. So I look at that as a great opportunity when we model our share of those new architectures, we have a great opportunity to grow our share. We have new capabilities that will expand our available market as those new architectures are adopted. And I don't want to give a specific forecast relative to timing. But what I would say is that we will see a pretty significant growth in next year. in our panel revenue. And certainly, it's going to ramp a fair amount after that going forward. But our position there are stronger. And then relative to hybrid bonding, hybrid bonding is a way that every every customer, whether it's leading edge, [ oundrylogic ] or DRAM or high bandwidth memory, they're all wanting to shorten the wiring length to improve the performance and power. So it's a very important inflection for all of our customers. Again, Applied has strong technology in hybrid bonding. We have also very big business in the adjacent steps around the hybrid bonding, we have the only integrated R&D facility, working with our customers to enable those new architectures for packaging. So this will be a meaningful growth driver over time in addition to all the other technologies that we have in advanced paging. So I'm very optimistic over 70% growth this year and very strong growth in '27 and going forward.

Then one quick follow-up for Brice. Brice, I understand you don't want to give us F Q1 guidance, but you did call out that it's a 14-week quarter. I'm just wondering what sort of impact, if any, that will have on your revenue sequentially.

Sure. What we've seen in prior years when we had a 14th week in the quarter was you get close to a ratable performance on the services side of the business and not so much on the equipment side, most of the planning is done at a quarterly basis on the equipment side. So yes. And then from a spending perspective, as you might imagine, everybody wants to be paid. And so most of the spending will occur in the quarter. Thanks, Srini.

So thank you, Srini, for your questions. And Brice, would you like to get a little summary before we close the call.

Thanks, Mike. We're excited that Applied's unique portfolio and strategy are enabling us to grow faster than the market and increase margins. We look forward to an even stronger second half and to seeing many of you at our upcoming events. I'll be attending Citi conference in New York, and Gary will be at the Goldman Conference in San Francisco. The whole team looks forward to seeing you at the EPIC Center in October and giving you our longer-term outlook at our investor breakfast at SEMICON West. Mike, please go ahead and close the call.

All right. Well, thank you, Brice, and we'd like to thank everybody for joining us today. A replay of today's call is going to be available on the IR page of our website by 5:00 Pacific Time, and we'd really like to thank you for your continued interest in Applied Materials.

Thank you, ladies and gentlemen, for your participation in today's conference. This does conclude the program. You may now disconnect. Good day.

Hello, and welcome back to the Applied Materials Master Class series. Several years ago, we anticipated that the AI wave would drive the semiconductor industry to $1 trillion in annual sales by 2030. And through that time, we modeled the wafer fab equipment market composition to be around 1/3 leading-edge, foundry-logic, 1/3 ICAPS and 1/3 memory, with memory evenly divided between DRAM and NAND.

Today, we see AI driving the semiconductor industry to around $1 trillion this year and new incremental applications like agentic, edge and physical AI growing the industry to much higher levels over the next several years. These AI waves are fueling demand for faster, more energy-efficient chips and systems, and this is creating a new WFE spending mix.

We now expect leading-edge, foundry-logic to outgrow ICAPS and drive well over 50% of foundry-logic in the years ahead. And in memory, we expect DRAM WFE spending to be well over 2x NAND spending. Applied is well positioned for this new mix with the highest process equipment market share in leading-edge, foundry-logic, which we covered in our April master class as well as both DRAM and advanced packaging, which we're covering today.

In a moment, Kevin Moraes will summarize our strategy and explain why DRAM and advanced packaging are growing with AI. Next, Sony Varghese will share the road map for both standard DRAMs and high-bandwidth memories. Then Jinho An will discuss how advanced packaging is enabling faster and more energy-efficient AI chips and systems. Next, Lior Engel will explain how we are bringing eBeam process control to advanced packaging. Finally, Kevin will give examples of how inflection-focused innovation is growing our business, and then the 5 of us will address your questions.

Now, Kevin, it's over to you.

Thank you, Mike. Applied's strategy is inflection-focused innovation. We collaborate with our customers to predict road map inflections and develop solutions to the highest value technology challenges. We do this using a broad portfolio of unit process systems, co-optimized solutions and integrated solutions. In our previous Master Class on logic, we showed you how our strategy has given Applied the #1 materials engineering share position in gate-all-around in both transistors and wiring with a combined share of around 50%.

I'll next share 3 examples of how we used our strategy to become #1 in DRAM and advanced packaging as well. Back in 2013, Applied's DRAM process equipment share was below 15%, and we've since gained around 10 points of share and become #1. In our first Master Class 5 years ago, we explained 2 DRAM inflections. The first was patterning, where we co-optimized our Draco CVD hard mask patterning film, our Sym3 etch and PROVision eBeam metrology to help customers scale DRAM capacitors. Today, these products are used by all our customers and generate over $0.5 billion in annual sales.

Another inflection was transitioning the DRAM periphery logic transistors from polysilicon to high-k metal gate and incorporating more advanced wiring technologies to increase DRAM performance and energy efficiency. This change has increased our DRAM revenue by more than $1 billion per year.

A third inflection was high-bandwidth memory. We formed the packaging group that combined our materials deposition and removal technologies to enable through silicon vias and microbumps. High bandwidth memory took off, and we've become #1 in HBM packaging and equipment sales. In fact, HBM is growing rapidly this year and helping expand our overall advanced packaging business by over 50%.

In today's Master Class, you'll learn about 5 emerging DRAM inflections and 2 advanced packaging inflections. In 6F2 DRAM, we see customers increasing the use of EUV patterning to further scale DRAM memory cells and peripheral logic. Next, we see DRAM customers following the foundry-logic road map adopting more advanced transistors and wiring in the peripheral logic to improve performance and power. In addition, we see a new CMOS-bonded array architecture in which memory and peripheral logic are built on separate wafers that are later bonded together.

Next, we see a new 4F2 DRAM architecture based on vertical cell transistors. And finally, we see 3D DRAM built using the same deposition and etch technologies that Applied leads in today. In advanced packaging, we see the industry moving from silicon wafer interposer substrates to panels, which enable larger body AI accelerators. We also see the eBeam technologies used in the leading-edge logic foundry and DRAM being adopted in advanced packaging.

Now before Sony, Jinho and Lior discuss these inflections, I'll provide a nontechnical overview of some of the key ingredients inside the computer to help explain the unique and critical role DRAM and advanced packaging play in the AI era. Regardless of the workload, computers need processors to execute software instructions and manipulate data. CPUs, GPUs and TPUs can process extraordinary amounts of data each second. Keeping them fueled with data to avoid idle states is imperative to performance and performance per watt. Since processors have limited space for memory, we prefer to keep data that is not actively being processed nearby rather than on chip.

The perfect memory would have low latency, high bandwidth, random accessibility, unlimited endurance and nonvolatility as well as low cost. But the perfect memory doesn't exist. And over the decades, a number of memories with different characteristics have been invented and refined.

Let's take a closer look at 3 key memories. SRAM is built using logic processor technology and can be integrated onto the processors, enabling cache memories with 0 wait state performance. SRAM is extremely fast. The downside is SRAM is typically requires 6 transistors for every bit of data stored which makes it particularly expensive. And after decades of optimization, designers are finding it increasingly difficult to scale SRAM more than marginally with every new node. SRAM caches can also easily take up 1/3 or more of the processor's die area and using advanced packaging to relocate these caches without sacrificing performance is a major focus of the industry today.

Next comes DRAM. DRAM processor technology is mostly incompatible with logic technology and resides off chip. It shares the random access capability of SRAM, but needs only one transistor per bit of storage, making it much more space efficient and cost effective. Each DRAM cell has a transistor that controls the binary state of the bit and a capacitor that holds a charge level that represents 0s and 1s. Each cell is connected to a word line and a bit line, which together enable a random access, read and write. Zooming out, each DRAM chip has gigabytes of memory cells along with the periphery logic regions that route data to and from the memory cell and to the rest of the computer system.

Finally, NAND. Each NAND cell includes a transistor that traps electrons and the resulting threshold voltage represents 0s and 1s. NAND memory cells do not have the space overhead of unique bit line and word line connections. Instead, they are arranged in long, narrow strings that offer higher density and lower cost versus DRAM. A special attribute of NAND is that the transistors can be programmed as a multilevel cell, whereby 4, 8 or 16 distinct voltage levels are used logically to represent 2, 3 or 4 bits of data in each cell.

The transition from single-level cell NAND to these multilevel cell variants has dramatically increased NAND storage density relative to DRAM. In fact, it can take nearly 200 HBM wafers to hold the data stored on a single QLC NAND wafer. But density and cost benefits of NAND come at the expense of performance and endurance. Strings of NAND cells are organized into large pages and blocks. Individual bits are not randomly accessible. Instead, an entire page of data, typically several kilobytes must be read to access any particular bit. Write operations occur at the page level, and erased operations can only be done at the block level, wherein millions of NAND cells are erased at the same time.

NAND chips constantly move fresh data to empty cells, so blocks of stale data can be erased. Erasing NAND requires more than 20 volts, and the operation degrades NAND reliability, leading to endurance limits. All of this helps explain why NAND is not a perfect substitute for DRAM and why designers instead use NAND more like a fast hard disk drive for data that seldom changes.

This table summarizes memory performance as measured in access time and energy per bit transferred. If you normalize accessing a bit of data from SRAM to 1 second, then accessing the same bit from HBM takes around 10 seconds and from DRAM module about 1 to 2 minutes. And from a NAND SSD, tens of hours. The energy consumption per bit transfer also multiplies substantially as data moves further away from the processor, from on-chip SRAM to high-bandwidth memory to DRAM modules and NAND SSD.

Returning to the overall system architecture, a key challenge today is what we call the memory wall. It describes the performance gap between increasingly fast processors and memory, which struggles to keep the processors busy. In AI inferencing, memory access and data movement can account for a vast majority of the overall system activity and energy use.

Next, I'll relate what we've discussed about memory to AI and what's called KV cache. In AI workloads, items to be processed such as language are divided into small segments called tokens that can be manipulated by AI models. Key value pairs are mathematical representations of tokens that are stored as the model works to understand context.

In Agentic AI, context windows grow longer and sessions persist across increasing numbers of steps, causing KV cache data to expand. DRAM is extremely fungible and preferred for performance-critical KV cache data that is frequently updated.

NAND provides a cost-effective repository for overflow KV cache data that seldom changes. Bringing together everything we've discussed, a key insight is that the closer we put memory of all kinds to the processor, the faster the system and lower the power. And this explains the opportunity to scale the entire system using advanced packaging techniques we'll describe today.

I hope my presentation helps you understand why we believe leading edge, foundry-logic, DRAM and advanced packaging will be an attractive area of the WFE market for as long as AI continues to proliferate and grow. It should also help you understand why we now expect DRAM WFE market to be over twice the size of NAND market for the foreseeable future.

And now I'd like to turn the meeting over to Sony Varghese, who will discuss the DRAM road map. Sony?

Thank you, Kevin. I've spent over 20 years in the DRAM industry, working at one of our customers for over a decade. There has never been a more exciting time for DRAM, both in the value DRAM brings to AI and in the innovations that's happening from the memory cells to the periphery transistors and wiring, and to the advanced packaging. I'll discuss 5 DRAM inflections that combine to improve DRAM density, performance and power consumption.

Let's start by defining the key DRAM structures. As Kevin mentioned, every DRAM cell needs a transistor and a capacitor. To hold a one in storage, the capacitor's charge needs to be refreshed around 15 times every second. You've heard about 6F2, which was preceded by 8F2 and will be followed by 4F2. F2 refers to the square of the smallest wire width in the DRAM cell, and the coefficient refers to the multiple of that area needed to construct a complete memory cell.

Within the memory cell, the capacitor acts like a tiny storage container and the transistor acts as a switch. The switch turns on to read or write data, and off to hold it in a place. Cells are organized in rows called word lines and columns called bit lines. When a word line is activated, all the cells in that row are turned on. Bit lines connect the cells to the surrounding control circuitry, which manages all key operations, reading, writing and refreshing the data so it's not lost. It also measures the electrical charges to determine whether each cell holds a 0 or 1.

As the memory cells shrink, they store less charge, making it harder to reliably distinguish zeros from ones. In NAND, sensing reliability challenges drove the industry from 2D shrinking to 3D stacking of large, more reliable storage elements. In DRAM, we still have opportunities to drive 2D scaling even further before we need 3D stacking. We see several more inflections in 6F2 DRAM in the coming years.

The first DRAM inflection is the greater use of EUV to increase DRAM density and capacity by scaling DRAM capacitor arrays, word lines, bit lines and shallow trench isolation. The EUV inflection is positive for Applied because our Sym3 Y Magnum system is the most widely adopted etch technology for EUV patterning in DRAM. The system acts on EUV photoresist, cyclically etching and redepositing the material in the same chamber to straighten lines that are uneven due to EUV stochastic errors. By making EUV line patterns smoother before they are etched into the wafer, we help increase yields and decrease line resistance to improve chip performance and power.

The second DRAM inflection is that the transistor and wiring technologies pioneered in leading-edge foundry-logic are now being used in DRAM. I'll share 3 examples. First, our customers are adding a new epitaxy step. Applied today introduced a Centura Prime Epitaxy system that has been specially designed for DRAM fabs. We selectively grow boron-doped, embedded silicon germanium in the source drain regions of DRAM periphery transistors. The resulting channel strain improves DRAM performance and power consumption.

A unique feature of the new system is a 20% smaller footprint, which helps DRAM makers generate more output per square meter of clean room. Second, as continued DRAM scaling with EUV packs more DRAM bits and periphery transistors on each chip, designers are adding more copper interconnect layers. Just as in foundry-logic, these additional layers increase the market for Applied's leadership products in wiring, including PVD and CMP. Third, DRAM makers are preparing to transition from high-k metal gate periphery logic transistors to FinFETs.

Just as in foundry-logic, FinFET transistors offer better control of the silicon channel, which increases performance and reduces power consumption. FinFETs also increase DRAM scaling in 2 ways. First, FinFET transistors are more space efficient than high-k metal gate transistors. Second, FinFETs have lower transistor to transistor variability. I described how the periphery logic senses the capacitor charges that represent zeros and ones and how the sensing margins are becoming narrower with scaling.

FinFETs allow us to continue to scale density and maintain reliability. The third DRAM inflection is the CMOS-bonded array. DRAM makers can build memory arrays on one wafer, periphery logic on the second wafer and bond them together. The approach lends itself to 6F2, 4F2 and 3D DRAM as well.

There are several benefits to decoupling logic from memory. One is density, placing memory and logic on different wafers allows each to be scaled more aggressively. A second benefit is transistor performance. The high temperatures needed for DRAM capacitor fabrication deteriorate FinFET transistor performance, which is maximized as the array steps move to a different wafer. A third benefit is interconnect performance.

In wafer-to-wafer bonding, the connections between memory cells and peripheral logic are shorter and more vertical. Designers can eliminate many of the long, horizontal connections that reduce speeds and increase capacitance, power and heat.

From a business perspective, the wafer bonding also adds new wiring steps that Applied can serve. The fourth DRAM inflection is a space-efficient 4F2 architecture in which vertical transistors are deployed in the memory cells to further scale both the transistors and the capacitors. I'll describe significant 4F2 materials engineering challenges and the products Applied is delivering to solve them.

The first is etching the vertical cell transistor's narrow, high aspect ratio silicon channels. The shapes need to be exactly as intended and the billions of channels across the chip on the wafer need to be uniform. Applied's Sym3 Z conductor etch system has been optimized for this application. A second challenge is managing the materials interface between the channel and gate oxide to reduce electron leakage, improve data retention, support capacitor refresh operations and minimize standby power.

Applied delivers 3 helpful steps. First, Applied's Centura DPX solution uses plasma to densify the gate oxide to improve its quality. Next, our Centura RPO2 Radiance system engineers the gate oxide interface to reduce the defects that trap electrons and holes. Finally, Centura DPX builds a thin nitride film barrier that promotes the adhesion of the subsequent transistor metal gate layers and prevents metal impurities from diffusing into the gate oxide. A third challenge is creating the transistor's multilayer metal gate stack.

The Trillium ALD system launched during our Logic Master Class in April is also being used in vertical cell DRAM to enable uniform metal gate stack deposition. A fourth challenge is reducing the area occupied by the metal gate stack to maintain the tight 4F2 design parameters. Etching is used to resist or bury the metal gate within a preformed trench in the silicon substrate. Applied's selective etch technology tightly controls the recess shapes, depth and uniformity. A fifth and particularly important challenge is forming the bit line junction to the cell transistor. A region of the silicon channel is doped with phosphorus ions, creating an electrical connection between the 2. This step is the first to require low-temperature selective epitaxy in DRAM.

In 6F2 DRAM, the cell transistors are U-shaped. The word line is placed within the U, and the bit line connections are made at the top of the structure where both ends of the transistor channel are accessible. But the 4F2 transistors are I-shaped, making it difficult to dope the lower end of the silicon channel. The junctions are deep and narrow and reliability depends on keeping the dopants contained to the target regions of the transistor and applied uniformly across all the transistors on the wafer.

In 4F2 DRAM, traditional high-temperature anneal processes are not viable because they damage the delicate capacitor structures. To solve this, Applied has developed a low-temperature selective epitaxy that grows doped crystalline silicon only on the targeted region of the transistor channel.

From a road map perspective, we see multiple nodes of 6F2 innovation before the arrival of 4F2 and 3D DRAM. However, R&D path finding in 3D DRAM is happening today. The work includes identifying the ideal materials and structures and learning how to deposit, modify and remove them to create reliable high-performance DRAMs that are scalable to much greater heights over time.

A key insight from the earlier portion of today's class is that 3D DRAM will be very different from 3D NAND, demanding a unique set of materials and materials engineering capabilities that are already leadership areas for Applied today. Earlier, we noted that DRAM is a random access, read-write memory that is heavily used in computing to deliver high-performance, low power and high endurance. In contrast, NAND is better suited to reading and writing blocks of data that rarely changes.

3D NAND is built by depositing alternating layers of insulating dielectric material and the memory holes are formed using dielectric etch. In contrast, 3D DRAM stacks will rely heavily on conductor and metal stack deposition and lateral conductor etching. Applied's leadership in epitaxy, conductor etch, eBeam metrology and inspection and other technologies give us a head start in collaborating with customers to develop 3D DRAM prototypes. In fact, 3D DRAM is one of the technology inflections targeted for customer co-innovation at the EPIC Center in Silicon Valley. Because we are doing pathfinding work today, we already have line of sight to many of the key process steps and high-value problems we can address with our products.

I'll discuss some of the major 3D DRAM materials engineering challenges and the solutions that Applied is providing. 3D DRAM requires building an ultra-tall stack of more than 100 alternating layers of silicon, silicon germanium. The superlattice is formed using advanced epitaxy. Maintaining tier-to-tier uniformity across hundreds of horizontal layers is essential, since variations impact electrical performance.

A major materials engineering challenge is managing interfacial stress within the silicon germanium lattice, which has a higher lattice constant than silicon. Without control, lattice mismatches can generate defects that increase DRAM cell transistor leakage and variability.

Applied is using its epitaxy leadership to tune surface chemistries and process conditions that help customers achieve high film quality and uniformity. Another challenge is structural integrity and uniformity of tall and complex 3D DRAM material stacks. The technology we use to treat silicon nanosheets in gate-all-around transistors can help strengthen the silicon channels in 3D DRAM as well. The 3D DRAM layers will be highly sensitive to impurities and particles.

Applied's integrated material systems can be used to maintain an ultra-high vacuum clean deposition environment. Once the 3D DRAM stack is formed, it must be etched at an ultra-high aspect ratio of more than 200:1. The etch pattern is defined from above. A CVD hard mask is deposited on top of the stack, patterned and opened with a hard mask etch step to create a stencil for a second etch step that will transfer the pattern through the stack below.

Next, Applied's Sym3 conductor etch system is well suited to etching the tall materials stack, providing excellent uniformity and depth control through hardware innovations like high-voltage bias pulsing and ion directionality. After the deep etch, a high-quality dielectric gap fill is needed to provide electrical isolation and structural integrity. ALD oxide is a good candidate for void-free gap fill.

Next, precise lateral recess etching is required to selectively remove sacrificial silicon germanium layers and define where subsequent structures like the metal word line will be formed. Even nanometer scale variations in lateral recess etch can degrade transistor performance and cause leakage.

A highly selective recess etch is needed to remove all the silicon germanium while leaving the adjacent materials and structures undisturbed, and Applied has industry-leading selective etch technology. The recess step sets the geometry of the word line that will be deposited within the tall 3D DRAM stack. A metal such as moly will be used for the word line. And as stacks grow to 200 layers or more, managing word line sheet resistance will be critical for performance, power consumption and switching uniformity across layers. One way to reduce resistance is by growing large, well-formed grains of moly that reduce electron scattering.

Applied's Viva High Power solution uses a radical and thermal treatment to precisely control grain growth. Next, selective epitaxy is used to form the lateral transistor junctions and storage node contacts. This epitaxy step is especially challenging because it is non-line of sight and must be uniform across hundreds of layers. Applied's Centura Prime Epi system is tuned for highly selective deposition and leaves no unwanted metal that could create shorts in buried structures that cannot be accessed or repaired.

Finally, selective ALD deposition of a metal such as molybdenum silicide forms low resistance contacts across all the layers. Lateral selective deposition is challenging and plasma energy cannot be used because ions travel directionally. So thermal processes must be used instead. And the selective ALD moly technology we developed for foundry-logic market can be used in this 3D DRAM application. I'll conclude my section with an overview of high-bandwidth memory.

HBM DRAMs are like standard DRAMs in most areas of wafer processing technology. The most significant design difference is the TSV channel holes that are large and consuming as much die area as 2,000 memory bits. As a result, HBM dies can be twice the size of standard DRAM dies, and the industry needs to produce 3 or 4 HBM DRAM wafers to deliver the same number of bits. Comparing HBM4 to HBM3, major differences include a much wider 2048-bit interface, a doubling in channel count and a tighter TSV pitch.

At the base of every HBM stack sits a logic die that helps transfer data between the HBM chips and the processor memory controller. In earlier HBM generations, the base die performed relatively simple control functions and was built on non-leading-edge logic. In HBM4, the base die has more transistors to deliver higher data bandwidth, improve signal integrity and enable power management. The HBM4 base die is moving to more advanced logic nodes with FinFET transistors and advanced wiring, which are better suited to running these more complex functions at a lower power. As a result, the growth in HBM demand increases Applied's foundry-logic business as well.

My colleague, Jinho will explain how TSV DRAM wafers are later processed to create and connect the through-silicon vias. Thank you for listening. And now Jinho, it's over to you.

Thanks, Sony. I worked at the world's largest memory company on technologies, including HBM before joining Applied 5 years ago to develop new advanced packaging platforms. At that time, we held our first advanced packaging master class and proposed that the future of chiplets that Gordon Moore predicted back in '65 was going to be a key driver of the industry's progress in the AI era. Today, no one doubts that packaging has become as strategic to the computing industry as on-chip scaling of transistors and wiring.

Let's take a moment to revisit how chip design was done when Moore's Law was working well, in the system-on-a-chip era. On the left is a server chip with a variety of functional blocks. With each new node, designers added more performance-enhancing transistors and on-chip SRAM memory. They also integrated more system functions that were previously handled off-chip on application-specific ICs.

Looking at the chip in cross-section, the tiny transistors were connected to successive layers of wiring as we saw in our previous Master Class. The more the transistors, the more layers of wiring. And the farther from the transistors, the thicker the wires and fewer the connections.

Looking at a modern AI super chip, you can see that as the pace of Moore's Law slowed, designers adopted a system-in-a-package strategy. Each processor approaches the reticle limit or the roughly 860 square millimeter size that an EUV litho scanner can print. And to keep up with growing AI performance and data needs, there are now multiple processors in the design. Some system functions have moved from on-chip to off-chip.

Looking at the system in a package in cross-section, we see many layers of wiring like in the system on a chip, but significantly more complex. I'll next aim to make all of these easy to recognize and understand.

To define the elements, let's begin with a look back at packaging and system integration during the system-on-a-chip era. This is a server chip and printed circuit board from 2016. Looking at a cross-section, we see that the CPU die sat on a chip substrate and the chip substrate was connected to an epoxy printed circuit board. This motherboard had wiring traces that connected the processor to DRAM modules, which each had a number of similarly mounted DRAM chips.

Even in today's most advanced packages, we still combine these 3 ingredients: dies, substrates and motherboards. But we had a fourth element, the interposer, which sits between the dies and the package substrate. The interposer is where much of today's advanced packaging innovation is focused. To understand the need for the interposer, recall that in on-chip wiring, the closer to the transistors, the thinner the wires and the more the connections.

In the system-in-a-package design, there are now far too many transistors and connections for any normal chip substrate to handle. The interposer has multiple layers of wiring called redistribution layers or RDLs. It acts like a miniature circuit board within the advanced package, enabling thousands of short, high-density connections between the multiple dies above and the substrate below. Compared to die-to-substrate packaging, interposers dramatically reduce interconnect length, which reduces resistance and power consumption while increasing bandwidth and performance. In short, the interposer marks the transition from packaging as a protector and connector to packaging as a critical enabler of energy-efficient performance.

The first interposers were made of silicon because the industry had decades of experience building layers of fine wiring on silicon wafers. Another approach is building layers of wiring into epoxy as an interposer material and embedding silicon bridges into the epoxy right beneath the dies. The bridge connects the dies to one another and the substrate below. The key benefit is cost as the epoxy substrates and silicon bridges cost less than silicon wafers.

This is a good time to define a few packaging terms. 2D refers to connecting package chips side by side on a circuit board, as in the older server example. 2.5D packaging is using interposers to connect multiple dies to one another on an intermediate substrate as we just discussed. 3D packaging is stacking dies on top of dies using through silicon vias and microbumps or hybrid bonds, which I'll soon describe. High-bandwidth memory is a great example of 3D stacking and 3.5D packaging is connecting 3D chip stacks on interposers or substrates.

Next, let's talk about packaging interconnects beginning with traditional 2D packaging. I/O and power circuits are often placed along the edges of chip designs. In simple chips, wires are bonded to these pads and the circuit board, so power can be brought to the chip and data can be routed between the chip and the rest of the system.

One of the original interconnect technologies is wire bonding, which has around 10 connections per square millimeter. In the '90s, as transistor counts increased, the industry brought RDLs to chip designs, thereby rearranging the chip's I/O connections as a rectangular grid pattern that covers the entire bottom of the chip surface with evenly spaced connection points. This change supports the use of flip chip packaging, which offers hundreds of I/O connections to the chip's substrate. Round solder bumps connect the substrate to the motherboard, and there are well under 100 bumps per square millimeter.

2.5D and 3D packaging use microbumps with I/O densities below 1,000 per square millimeter. 3D packaging typically relies on through silicon vias or TSVs, the outlines of which are defined during wafer fab production. After wafer processing and TSV fabrication, small copper pillars and microbumps can be used to connect the chips using a technique called thermal compression bonding or TCB with I/O density approaching 10,000 per square millimeter.

Hybrid bonding is the ultimate interconnect technology. It completely eliminates bumps and pillars. Instead, we align the on-chip copper wiring of 2 chips and use heat to fuse the copper of both chips. I/O densities now approaching 1 million per square millimeter are possible. An important observation is that the smaller the packaging interconnect, the lower the energy needed to transfer data, which is measured in picojoules per bit transfer.

In advanced packaging, bringing chips closer together and decreasing the interconnect pitch increases performance and reduces power consumption. Finally, co-packaged optics puts tiny fiber optic interconnections right next to the chip, so data moves as light instead of electricity, further increasing I/O bandwidth and energy efficiency.

I'll talk more about this later. Next, I'll take a moment to describe substrates, which holds a single chip or an interposer full of chips. A key function of the substrate is to support chips, remaining flat and rigid even as temperatures fluctuate. The substrate includes large RDLs that route power and signals from chips above to the circuit board below.

The RDLs are built one layer at a time, beginning with the first layer that is built on a sacrificial substrate that is later removed. Smaller bumps are added to the top of the substrates to connect the chips and the interposers, and larger bumps are connected to the back of the substrate to connect to the circuit board. There can be large differences in how the silicon chips, substrate materials and substrate wiring react physically to changes in heat.

Soft organic substrates are often infused with glass that adds structural integrity. The larger the substrate, the higher the risk of warpage, which is a key source of alignment and connection errors that can impact performance, yield and reliability. Lior will discuss the growing need for packaging process control.

Now I'll discuss a major emerging trend called panel, which will give the industry a larger interposer form factor that enables bigger AI superchips and higher manufacturing output. AI accelerator designers commonly describe the size of their products using reticle-equivalent area, which I defined earlier, and they hope to deliver 14x reticle size products within a few years. But silicon interposers are limited to the size of a 300-millimeter wafer minus the edge loss that comes from packing rectangular interposers on a round host. Only around 4 to 5 interposers of this 14x size can be made from a silicon wafer.

Fortunately, the packaging industry has long used much larger form factors for chip substrates. And today, we aim to enable panel interposers as large as 310x310 millimeters, 510x515 millimeters or even 600x600 millimeters. The bigger the panel, the more interposers that can be built at a time, increasing output and reducing cost.

Panels can be made of epoxy, advanced laminates or glass. Glass is an interesting material because silicon chips and certain types of glass expand at similar rates in the presence of heat. In addition, glass is exceptionally flat and rigid. It also has electrical insulation properties, which enables high signal integrity at high RDL wiring speeds.

Kevin described how Applied's pioneering work in TSV technology helped Applied become the #1 high-bandwidth memory equipment provider. Today, we are working years ahead of the panel inflection to build a comprehensive portfolio and help our customers and the AI ecosystem accelerate the panel inflection. Our portfolio includes digital lithography used to pattern RDL circuitry that exceeds reticle dimensions.

Our unique litho technology dynamically compensates for the topographical variations in large multilayer substrates. Around 5 years ago, we acquired panel PVD technology, and we've since expanded the platform to include CVD and etch. Our long heritage in the display industry helped us refine our large area CVD capability for panel, and we've ported our display eBeam review technology to panel as well.

Just recently, we acquired large area copper electroplating with the acquisition of NEXX. Applied's broad panel portfolio enables us to pioneer co-optimized solutions for panel manufacturing challenges. One example is through glass vias, which are challenging because copper and glass expand and contract at different rates, which can cause panel cracking and failure. Applied addresses this by tightly co-optimizing through glass via etch, CVD liner deposition, PVD seed layer deposition, plating, anneal and CMP steps. No other equipment company can do this.

Next, I'll explain how HBM memories are packaged. Earlier, I described that the spaces for the TSVs are patterned when the wafer is processed. Around 19 materials engineering steps are used to complete the TSVs and add metal pillars and microbumps that allow the HBM chips to be connected to one another in a thermo-compression bonding machine. Applied offers products for 15 of the steps, which occur on both sides of the wafer. Focusing on the front side, the first step is etch, which opens the spaces used for the vias.

Next, an Applied CVD system deposits a highly conformal dielectric barrier film that will electrically isolate the copper TSV from the silicon. Then an Applied CVD system deposits a tantalum nitride barrier that promotes better adhesion of the subsequent liner to the CVD dielectric, a tantalum liner that prevents copper from diffusing into silicon and a copper seed layer that enables good copper fill.

Next, an electroplating step deposits copper into the vias. Finally, an Applied CMP system removes excess copper and creates a perfectly flat surface for the next process steps. Once the TSVs are completed, microbump pillars are formed on the top of the wafer, which is then bonded to a temporary carrier and flipped upside down. The backside of the wafer is ground to remove bulk silicon and CMP is used to thin the wafer to just above the backside of the TSV. A highly selective recess etch exposes the TSVs without damaging adjacent structures.

Next, an Applied CVD system deposits a passivation film that electrically isolates the exposed TSVs and protects the bottom side of the chip from copper migration. We use a special low-temperature CVD film because high temperatures would damage the delicate TSV and DRAM features. Our CMP system is again used to make all the structures level and ensure a flat surface for the backside metal pad formation steps.

Applied's leadership in the most critical of these TSV formation and interconnect steps has given us the #1 market position in HBM packaging. Applied today introduced a new product that helps our customers work with thinner HBM DRAMs to squeeze more of them into a given stack height. HBM dies are thinned to around 1/25th of the height of a standard DRAM wafer, which makes them susceptible to bowing. The higher the stack, the higher the risk of alignment errors, bonding failures and yield issues.

Our new Avila 2 CVD system uses a multistep sequence to increase the rigidity and reliability of HBM dies. It first deposits a thin silicon nitride layer that electrically isolates exposed TSVs on the backside of the wafer. The step balances front side film stress to help prevent bowing.

Next, the system deposits a thicker oxide film that further stabilizes the chip, creating a robust, uniform surface for the subsequent CMP polishing step. Avila 2 enables reliable stacking of 12, 16 or even more thin HBM dies.

Next, I'll discuss 3D chip stacking with hybrid bonding. Hybrid refers to the fact that both the copper wires and the dielectrics surrounding the wires of 2 chips are connected in the same operation. The first process step is CVD dielectric film deposition, which electrically isolates the copper wires of the 2 chips. Next, etch defines spaces for copper bond pads, which are aligned with each chip's copper wiring.

Next, we use an integrated system that uses PVD to deposit a thin barrier liner that prevents copper from diffusing into the dielectric material and promotes good adhesion to the CVD film. Next, PVD deposits a thin copper seed layer that promotes high-quality copper film. An electroplating step fills the lined bond pad trenches with copper.

Next comes one of the most important steps in hybrid bonding. CMP removes excess copper and makes the copper wires and surrounding dielectrics perfectly flat and uniform. Applied's Opta CMP system dishes the copper wires, creating nanoscale divots in the wiring, so it is recessed when the 2 chips are later bonded together.

Next, a plasma preparation step activates the dielectric films for chemical adhesion and also cleans the 2 surfaces. Next, the chips are perfectly aligned and pressed together in a hybrid bonding machine. An anneal step strengthens the dielectric film bonds and expands the copper wiring so it fuses together.

Applied has introduced Kinex, the industry's first integrated die-to-wafer hybrid bonding system. Kinex integrates plasma surface activation and cleaning, bonding and metrology in one controlled flow. Integrated hybrid bonding reduces the queue time between the critical activation and bonding steps, which preserves bonding integrity and improves bond strength and yield. The enclosed system minimizes particle contamination, which is critical because hybrid bonding uses known good dies and needs to be nearly 100% defect-free.

Our integrated metrology makes sure the chips are perfectly aligned. Kinex is designed to help expand die-to-wafer hybrid bonding from a high-end niche application to the mainstream. I mentioned how hybrid bonding consists of several CMP steps. Compared to wafer processing CMP, advanced packaging applications introduce new challenges, including thicker films, longer polish times and tighter tolerances. Longer polishing times can cause variations from the perfectly level and smooth surfaces chip makers need in applications like hybrid bonding, where even tiny variations in wafer center-to-edge planarization can cause expensive yield losses.

Opta Quad has a unique, real-time process control system that actively monitors and dynamically adjusts the planarization process on these long-duration, multistep advanced packaging CMP flows. Today, Applied introduced Nokota VMax 2, which is a next-generation electrochemical deposition system purpose-built for advanced packaging applications like HBM.

Plating is a foundational step across interconnect schemes, from TSVs within dies to microbumps that connect stacked dies. As TSV dimensions shrink, it becomes increasingly difficult to deliver chemistry deep into the narrow futures while maintaining high-quality fill. Nokota VMax 2 is optimized for void-free, defect-free metal fill even below 3 microns.

HBM stacking also raises the bar for plating tools, requiring tighter bump coplanarity and higher throughput. A key challenge is pattern density variation across the wafer, which distorts current distribution and leads to nonuniform metal deposition. Nokota VMax 2 addresses this with Adaptive Pattern Tuning, dynamically shaping the electric field to correct for layout-driven variation and improve uniformity and coplanarity. This control extends to both sides of each HBM layer, which is critical in multipass plating flows.

My final topic today is co-packaged optics or CPO. Optical fibers have long been used to transmit digital data at high speeds over longer distances, such as between racks and even between data centers. In co-packaged optics, electrical signals are converted to light waves that are transmitted through optical fibers. Co-packaged optics is enabled by advanced packaging technologies, including hybrid bonding that brings electronic and photonic chips close together so they can operate as a single, tightly integrated system.

Optical modulators convert electron signals to light signals. Light generating devices such as lasers are integrated close to or even on silicon chips. Photonic waveguides can guide light across the chips, and they are fabricated by depositing a silicon or silicon nitride layer that is etched to form narrow features that light can travel along. Fiber couplers transfer light between photonic chips to external optical fibers.

To help develop the optical I/O ecosystem, Applied Ventures is partnering with optics and photonic start-ups along with manufacturers, including GlobalFoundries. We also have R&D programs in the materials, process technologies and equipment needed to produce and integrate optical components. Our equipment opportunities include epitaxy, CVD, PVD, etch, CMP and process control.

Now I'll summarize by revisiting what I said at the beginning that the future of chiplets that Gordon Moore predicted will be a key driver of AI today. If we had stayed with system on a chip and Moore's Law scaling, then today's server chips would have around 200 billion transistors. Thanks to advanced packaging, today's AI server chips have over 1.6 trillion transistors and pack 4x the silicon content in every socket. By going beyond Moore's Law, we're enabling new applications like training, inferencing and agentic AI that depend on advanced packaging for performance, power and cost.

Now I'll hand the meeting over to Lior Engel to discuss packaging process control. Lior?

Thank you, Jinho. The packaging market has long been served by low-cost optical and physical process control tools, but particles that were acceptable when large interconnects and bumps were used are turning into yield killers in hybrid bonding applications. Structures that could be easily measured by using optical tools are becoming so small that eBeam resolution is needed and delicate wiring can be damaged by wafer probers. So as packaging advanced from 2D to 2.5D and 3D integration of known-good dies, packaging fabs increasingly need the same advanced process control technologies found in wafer fabs.

I will briefly describe 4 key benefits of eBeam in advanced packaging. One, eBeam columns produce a finer beam than optical tools. While optical systems struggle to image tiny defects and submicron features, eBeam can image single-digit nanometer features and measure them with sub-nanometer precision. Second, eBeam can be tilted to profile sidewall and measure high aspect ratio structures. Third, eBeam enables materials identification, which is critical to deciding the corrective actions to take when yield issues arise. And fourth, eBeam can test the electrical functionality of delicate circuitry without making physical contact. These attributes make eBeam valuable in advanced packaging where proper manufacturing of features like TSV, microbumps and hybrid bonding pads is critical to chip performance, power and yield.

Applied has been in the eBeam business since the 1980s. And today, we have the #1 market position. We are now transferring our eBeam expertise to the advanced packaging market with new products that will help our customers solve critical challenges. Jinho's presentation has given us insight into how eBeam technology can help. For example, in stacking SRAM cache memory chips on top of processors, also in stacking HBM chips, transitioning from 12-high design to 16 and beyond, and in TSV critical dimension metrology to ensure full HBM stack connectivity.

Another challenge in advanced packaging is handling the wide range of substrate Jinho described. This includes new materials like glass as well as geometric consideration, including thickness and warpage. Today, we are introducing 2 new eBeam products for our advanced packaging customers. Both were developed to bring the full benefit of eBeam technology to a wide variety of substrates, geometries and materials from silicon to organic and glass.

Applied VeritySEM eBeam system is used in wafer fabs to measure critical dimensions such as height, width and slope of gate-all-around transistor structures. Our new VeritySEM 7AP is designed for advanced packaging applications, including measuring hybrid bonding pads, TSV dimensions and microbump geometries. The system automatically configure itself to work with different substrates, materials and geometries. It works in line, delivering precise 2D, 3D and high aspect ratio measurements, including a unique tilted view of height and sidewall angle measurements.

Next, Applied SEMVision is the #1 eBeam defect review system in wafer fabs with over 70% share of the global market. Our new SEMVision G7AP is designed for advanced packaging fabs, automatically handle the full variety of substrates. Customers use optical tools to quickly scan for potential yield-killing defects, which are often too small for the optical technology to identify clearly. Our new SEMVision G7AP uses the superior resolution of eBeam to clearly identify and automatically classify those defects, enabling customers to rapidly address the root cause of advanced packaging yield issues.

VeritySEM 7AP and SEMVision G7AP are already in production at multiple leading advanced packaging customers in both memory and logic. Applied has additional advanced packaging process control systems in development, including optical and X-ray products that we will discuss in the months ahead.

And now I will turn the meeting back over to Kevin Moraes. Kevin?

Thank you, Lior. In a few minutes, we'll begin the Q&A. But first, I'd like to share a few examples of how our inflection-focused innovation strategy is driving growth for Applied in both DRAM and advanced packaging. Back in our 2021 Memory Master Class, we projected that co-optimizing our Draco CVD patterning films, Sym3 etch and PROVision eBeam technology could give us $1 billion of cumulative revenue opportunity between 2020 and 2024. In fact, we delivered more than $1 billion in sales from these products over that time. And today, we believe we can generate more than 4x that over the next 5 years.

We also projected that improving DRAM periphery transistors and wiring with foundry-logic technologies could give us $2 billion of cumulative opportunity from 2020 through 2024. We delivered more than $2 billion from these applications over that period, and we believe we can generate more than 3x that over the next 5 years.

Turning to advanced packaging. We projected in 2021 that our business would more than double between 2020 and 2024 to over $1 billion. In fact, it more than tripled. And this year, we expect our advanced packaging revenue to increase by more than 50% to more than $2 billion. We continue to use our inflection-focused innovation strategy to predict road map inflections and deliver new products that enable our customers to accelerate their road maps.

We hope to see many of you at the unveiling of the new EPIC Center in Silicon Valley in October. At EPIC, we will take inflection-focused innovations to a new level by co-innovating with our customers in the industry's largest collaborative research facility. Many of our leading customers and partners have already announced plans to join us at EPIC, and there will be more announcements in the months ahead. In addition, we are completing the construction of a new EPIC advanced packaging center in Singapore. Now it's time for me to join the rest of our team for our Q&A session.

Welcome to the Bank of America Global Technology Conference, Vivek Arya from Bank of America, Semiconductor and Semicap equipment team. I'm really delighted to kick off the semiconductor part of our Global Tech Conference with the team from Applied Materials. I'm really delighted to have Brice Hill, the Chief Financial Officer. Warm welcome, Brice.

Good morning. Thank you for having us, Vivek. Glad to be here.

Yes. Thank you so much for joining us again. And I'll go through my questions, but please feel free to raise your hand if you would like to bring up something on your own. So, Brice, I guess, understatement that very exciting times in semiconductors. Every day you wake up, there's like some more news about demand growing. Maybe just walk us through your state of the union, right, at the start of the year, Applied was looking at a growth rate of 20% in your semi systems business, you upped it to 30%. Where have been kind of the incremental growth drivers in the near term, and then we'll talk about what you're seeing on -- over the horizon?

Sounds good, Vivek. So state of the union, the union is strong. We increased demand, and we increased our forecast, as you highlighted there to over 30% growth for our semiconductor systems business and that was an increase over prior quarter, as you saw new orders coming in from customers. And no surprise, at least relative to the news flow, we were just chatting about incremental investments that are being announced across the ecosystem. I think what's different about what's happening today is people can really connect the dots between the demand function of AI.

Our companies are using AI. We're deploying it in multiple ways. People are using it personally. So when they hear the cloud service providers talk about hundreds of billions of dollars of capacity adds to drive the compute behind AI, nobody is surprised. They can connect the dots to the demand. And that's certainly what we're seeing in our business. We talked about how leading-edge logic is growing very strongly. We talked about how DRAM is growing very strongly, and we talked about how advanced packaging, which is used to build the connected systems of all these components and drive the AI usage models, that's also growing very strongly in our business.

And those things, we received new orders in the last quarter. Customers are being more and more creative about finding clean room space and options to expand their output. That translates into new orders from us, and then translated into a higher growth expectation for this year. I know we'll talk about this, but we're -- the customers are working closely with us to give us a good perspective of what their growth plans are over the coming years.

Then in our services business, we had previously said, our services business would grow in the low double-digit growth range. I think the easy way to think about this is there's 2 drivers of growth in the services business. One is the installed base of equipment, and the other is the new products that we announced and create each year, which grows the value of each tool in your installed base. Because the systems business is shipping more and more tools, the demand we just talked about, the installed base for the services business grows and so we see that happening. We increased our expectation for growth in the services business to mid-teens. And we said actually this year will be a little bit higher than mid-teens because this year gets a plus also from the increasing utilization across the entire ecosystem.

So as the DRAM fabs went from whatever utilization they were at to 100% utilized. And as leading-edge logic goes to 100% utilized and even the ICAPS, the mature nodes, utilizations improve, that helps drive some uplift for the services business. So really a strong signal across all of our businesses as we look at the demand type today.

Got it. If we were to try and dissect this upgrade of the growth rate, does any specific areas stand out, Brice, like whether it was foundry logic or DRAM, NAND or advanced packaging, like did something really start to grow much faster than what you had thought before?

Well, I think in the short term, it's hard to separate leading-edge logic and DRAM. I think both of those are a very strong signal. And we didn't call one higher than the other. I guess as we look at factory projects that have been added over the last quarter, this is more longer-term signal. I think there's more DRAM adds than there are logic adds, but at least the way I think about it, and everybody can start to dissect the architectures of newly announced server systems and system platforms, but these things are traveling together. You've got a significant amount of leading-edge logic between GPUs, CPUs and accelerators paired with different memory types in a system. And so those things are being driven together.

Yes. Got it. And what are the main areas of constraints that you see? So theoretically, what can make this X percent growth rate more than X right now, why can't it be 35% or 40%? Are you limited just by clean rooms? Are you limited by your ability to produce new tools, the ability for the customers? Like, what are the main areas of constraints right now?

Yes. I think the primary -- we said that demand right now is metered by clean room space. And I think that is the right way to think about it. We've been talking about getting an 8-quarter rolling forecast from our customers, which we consolidate and give to our suppliers. And we've talked about, Vivek, how we added capacity, our own manufacturing capacity. So I think with the right signal, we can increase our output, and we're not the limiter from that perspective.

We certainly have manufacturing capacity available. And if we give our supply chain to lead, we think. So what's really the governor here, the governor here is it takes 3 or 4 years to build a fab. And so you need a pipeline of fab projects, which there is a pipeline of fab projects, but if you want that to go even faster, it takes some time to increase that pipeline. And I think that's what all the customers are working on. So in the short term, it's really clean room, making more clean room available and that's what -- in this year, that's what we saw some of the customers doing is even optimizing the clean room they do have available.

Got it. Yes. Actually, I wanted to pick up on that. When you say optimizing the clean room, are they upgrading to tools sooner than they would have otherwise? Like what does optimization mean to Applied?

Yes. I think we saw a number of different example types of literally just optimizing clean room floor space, so you might change layout, upgrading tools that were underperforming, upgrading process technologies from one node to another node. Those are some of the examples we saw as well as even some utilization of 200-millimeter space for 300-millimeter purposes. We see all of these things going on in the ecosystem.

Got you. And then you mentioned there is a 8-quarter kind of rolling visibility, what is your normal visibility? Like how should we put this 8 quarter in context of what Applied has seen historically?

Yes. It's great because first, I want to say that we have -- we always actually have a little bit longer than that. So we know just like people here know, because you see public announcements, we know all the fabs that are being built by our customers, although it's more precise for the largest customers than it is, for example, of some of the smaller customers in China, right? But you have an indication of what is out there 4 years from now, 5 years from now, 3 years from now, that goes into the fab tracker that we talk about and the like.

But when we start talking about 8 quarters, what we get from our customer when you go to 8 quarters is you get the precise process technology, the actual tool recipe that they're going to deploy, which tells us the types of tools, the number of tools and the specs of those tools that will be needed for that particular project as well as when they need to land, and we install those tools also, so we have to have install tools -- teams available to do that work. So it's really connecting the dots at a much more precise level.

And then when we get that from our customers, we consolidate that forecast, turn it into a forecast that we can provide to our suppliers, our 2,000 suppliers. So adding up all of -- if you think about adding up TSMC and Intel and Samsung, all of our large customers, and they all might need power supplies and you're aggregating that and saying, these are how many power supplies we need for the company, specific types and translating that to your supply chain so that they can put the capacity in place to deliver that product. That's really the purpose of getting that type of visibility. And I think it's working very well for us at this point.

Got it. So historically, the semi cap industry, and then CapEx has kind of grown 8% to 10% a year. And now the industry is at a place where you could have multiple years, conceptually, up 20%, right, 25% plus type of growth rate. What are you doing differently from an operational perspective to make sure that you're able to support customers for that kind of growth?

Yes. I think since COVID, Applied Materials, and I'm sure a lot of companies in the ecosystem, has been strengthening its manufacturing from a regional diversity perspective and also the supply chain from the same perspective. So we have more suppliers, we have more regional diversity. We have more regional diversity in our manufacturing and more raw capacity in our manufacturing. Those are definitely some of the things that we've been doing as well as working more closely with the customers than ever before on what their plans are, and what the scenarios are for the business.

So -- but I think, too, that since these are long lead investments when we were evaluating our own capacity additions over the past few years, since the business is profitable, you are sort of looking at what would be the max output required. And you don't want to miss capacity solutions if that max demand signal occurs. So you have to balance that in your planning, which comes to scenarios. Now did anyone forecast what's happening today exactly? No, but from a capacity planning perspective, you want to make those things possible.

Got it. Now specific to the growth rate for this year, how much of that do you think is share gains? How much of that is just kind of broader industry growth? Last year, there was a share shift, right? And just the same kind of moved against Applied. This year, do you think kind of the addressable opportunities moved towards you? So how much of this is share gain versus just broader industry?

Yes. This year, the mix -- I think, there's two things going on. This year, the mix is certainly friendly to Applied, as you said, having leading edges and DRAM and advanced packaging, where we have leading positions as process equipment suppliers, that's going into our wheelhouse, as you might say, but at the same time, even if that mix wasn't there, we expect to grow share in leading-edge logic, and we've talked for several years about how the transition to the gate-all-around transistor and to the newest wiring techniques will add share for Applied Materials since we've got solutions in those spaces.

We have been effective at growing share in DRAM over the last decade, and we expect to continue to grow share in DRAM. Some of the construction techniques used in leading-edge logic are being deployed in DRAM, and that's helping us add share there. And then in advanced packaging, we have a leadership position in advanced packaging, a lot of investment in advanced packaging. As we move from 2.5D, which is sort of lateral connection of chips to 3D when you're stacking chips, Applied has been investing in a lot of different packaging technologies that will be used in that roadmap as well as different substrates that may be used in that roadmap like panel, like glass, different substrates.

We've invested in hybrid bonding. We recently did an acquisition of an x-ray technology that will help us with evaluating the quality of those interconnects. We're in the process of another acquisition that helps us on the advanced packaging with fine-line interconnects between devices. So we're really focused on that portfolio as a huge opportunity for the industry and a big opportunity for Applied.

Got it. One last question on kind of the broader market, and then we'll go to the individual areas. Historically, when we just look at the ratio of wafer fab equipment to semiconductor industry sales, that's kind of varied between, say, 12%, 13% to as high as 17%, 18%, right? So the assumption has been that somewhere in the middle, right, 15% is a decent long-term number. But this year, a lot of semi industry growth is coming from ASPs rather than units. So do you think over the next 2, 3, 4 years that this will mix more towards units rather than ASPs is because units is really what will drive WFE, right? How do you kind of conceptually think about the density?

You're right. It's a great question because at least for me, I don't use that equation to think about it. But to your point, I think it was 14-ish percent last year, if you used last year's numbers. And it is subject to movement because of the different ASP trajectories that we see. But I like thinking about, as you're sort of suggesting actually that what's happening is more units are being required and more important for Applied Materials, more wafer starts are being required. So if you think about, for example, DRAM, there's 2.3, approximately 1 million wafer starts of DRAM every month. And when we look at the roadmap, we expect to see 300,000 or 400,000 wafer starts of added capacity every single year in DRAM.

So I think you sort of led in this direction, Vivek. The way we think about it is the actual wafer start output, the capacity of wafer starts is growing at the strongest rate it ever had -- has for DRAM, for leading-edge logic. And then the good news also is you have to connect those devices, so that will lead to packaging factories and packaging equipment that supports that.

So that indicator may or may not be the easy way of calculating equipment. It might be look at the amount of factories that are being built, look at the amount of wafer start increases that will go out, and that will give you an indication of the equipment business.

Got you. You mentioned DRAM, as many investors are very happy to see, right, the strong performance of DRAM and a lot of memory stocks, but there's also kind of this lingering concern that how much of this is right cyclical and -- so just theoretically, let's say if DRAM prices go down next year, do you think customers will be as motivated to add capacity? Or will they want to protect pricing? What has happened historically?

First, it's hard to picture that, but I'm open to scenario planning, so we can scenario plan. I think that one benefit the equipment side of the business has, is back to our discussion of what is metering the industry and what is currently the gate for the industry. I think it'd be very difficult to argue that we're in a position where we have too much capacity, and we've overbuilt capacity in almost a good way since it is metered, you're sort of behind the demand curve from a capacity perspective.

Certainly, if prices go down in the end devices that will -- all the customers will recalculate their ambitions and their interest in how much capacity to add. But right now, the profitability is so significant that I think there -- I think they're really trying to figure out how can they maximize output over the next 5 years.

Right. And I would assume that as many of them are signing long-term agreements with their customers, right? They're also on the hook to make sure that there is some capacity, right, to support all those?

Absolutely.

Right? Agreements. Staying on the topic of memory, how is your content evolved, right, from the various versions of HBM3, 4, and I saw last night, Samsung is now starting to talk about HBM5 as well. So how is your kind of opportunity and content involved in these different HBM generations?

Yes, I think the way to think about it for Applied Materials is, as you move forward on these HBM technologies. We talk about -- it increases the overall demand for DRAM. And it also -- there's a trade ratio on the HBM DRAM where it was around 3:1. So it takes 3x the wafer die area to make a DRAM device, if it's going to be stacked. And that trade ratio is moving to 4:1 now with the newer generations of HBM.

And so DRAM is already operating at full utilization. We're only adding more requirements on to DRAM. So I think this just ups that demand signal even more strongly than where it was before. And our customers, we expect, will be adding factories. Like we saw in the last quarter, they'll be adding factories and telling us shortly with our 8-quarter forecast that they'll need more equipment for those factories. And so it's just a -- since it's a great solution for the AI data center servers, and the best memory for those, then it's an added demand signal for us.

Got it. So I think Applied has mentioned many times that you gained 10 points of DRAM, WFE share, right, over the last decade or so. What do you think help to drive those share gains? And as you look at the next 3 to 5 years, what can help maintain or expand that market share?

Yes. I think it's -- what we've discussed is some of the same capabilities that we built in leading-edge logic on the transistor side are being deployed in DRAM from an equipment capability perspective. And that's really what's helped the share gain over the years, and I think that roadmap continues.

So we talk about periphery, IO, logic and other elements of the architecture that require those techniques. I think that's what is helping us win positions in the forward DRAM. And we've said that if companies move to 4F squared, that will be a share increase for Applied Materials, as we have a stronger position in the 4F squared than even in 6F squared. And when they go to 3D DRAM, same thing, we expect an uplift in our share position in those because of the long lead work we've been doing with the companies on the architecture and the equipment solutions.

Do you have some sense of timing of when we move to 4F squared or 3D DRAM?

For that, we just say we should -- you should talk to customers. So I think our customers, the roadmaps have been changing. They've been adding more 6F squared process upgrades in the recent. So I think each customer has a different expectation for when they would deploy those.

Got it. Okay. On advanced packaging expected to grow over 50%, right, even faster than the overall. So if you look over the next 2, 3 years, are there certain new emerging areas in advanced packaging, where you think Applied might be better positioned too to participate?

Well, definitely. So on the 3D stacking in general, Applied has technologies that we -- that will be deployed by customers, and then I mentioned a few of them a little bit before in highlighting that we're building a portfolio of solutions for packaging. So one acquisition that we're underway on right now, for example, is for fine-line interconnects. You can think of it, the substrate at the top of the substrate between chips that are being interconnected. The x-ray acquisition that we did is to examine the quality of interconnection of micro bumps between stacked chips and allow you to see things that we couldn't see before, the industry couldn't see before, and we've been working for years on hybrid bonding.

We've been working for years on panel technology. We have lithography technologies that we're deploying in different packaging solutions. So we really feel like we have a strong portfolio, and we're examining the evolution of larger and larger systems with more and more chips, which create all these different materials and equipment problems like warpage, like thermal management, hotspots, just the ability to get the signals on and off the chips and what technologies should be used for that. Applied Materials is investing in a portfolio that we think will be very strong as that roadmap continues to develop. And we think it's a huge growth area for the industry.

Got it. On the foundry-logic side, one very large customer, of course, has been driving most of the industry growth. Are you starting to see any sense of diversification, there's an emerging new player in the industry, right? Have you started to see anything new there? Like how do you look at the diversification over the next few years?

Definitely. We said in our earnings call, I think that we're shipping to 4 customers today on leading edge and that is becoming more and more diverse and robust. And of course, that's a great thing for the industry. I think competition in the foundry business will help drive the roadmap. And from an equipment and a factory perspective, it drives more solutions. So I think that's a good thing. And we do see interest from new customers.

Got you. Okay. On gross margins, Brice, there is always this kind of good problem to have where your customers are making so much more money, right, by way of gross margin. So what levers does Applied have to kind of exert your position in the industry, right, and expand gross margins, right, from -- I mean, you're getting towards 50% now. How much more upside is there?

Yes. Investors might notice that we've started reporting our gross margins by reportable operating segment. So you can see our equipment business has 54.8% gross margins, and that's improved quite a bit over the last couple of years. Vivek, I know we've talked about -- we've been working on our pricing process, building a robust, disciplined process where we set targets for every tool. We set -- we base it on our value analysis of what the tool is providing.

And so we feel like we've done a very good job of making that a more robust process for the company, and that's helped us with our margin uplift. And I would just say that in general, Applied Materials' strategy is to work with customers on the forward technology. So when we think about EPIC and our platform that we're building, the lab that we're building in Sunnyvale, where we've announced 10 customer arrangements at this point where we have some of the top universities and all the top customers partnering with us on that platform to work on -- collaboratively work on the furthest out technologies together. That's been our strategy.

And so what that helps you connect the dots as those portfolios are introduced, our gross margin, we expect will increase over time because that long-lead work that you're doing is on the most difficult technology problems for those customers. And so that's been our strategy, and we're increasing our capability of pursuing that strategy by building the EPIC platform and having the customers join us and working together. And the simple way to think about that is your core, your newest tools and your newest capabilities and your top researchers are all there. And if we can have the customers join us there and bring their architectures and their most difficult problems, then we expect to speed up progress in that way.

Got it. You raised the growth on your Applied Global Services, AGS, business towards mid-teens. Do you think there's a point at which you see the growth rates on the product side and the services side converge? Or this is just a long-term kind of secular grower and right and it's not -- it's countercyclical to some.

Well even when it was low double digits, at that point, it was probably ahead of where the equipment is. So we do think it's secular, if you will. We think the underlying drivers are an ever-increasing installed base and those tools tend to have -- be long-lived. So when you put new tools in place, those can be tools that are serviced between 10 and 30 years literally and so it's long life from that perspective. And then we continue to announce new products in the service space.

Most recently, a lot of these products have been AI-driven, where we're using AI to help make predictive recommendations on maintenance and upgrades to the tools that help the customers with output. And so as we make new service product announcements that grows the value of each tool in the service space. So we do think that is secular and goes back to the sort of the secular growth in the equipment itself.

Got it. One question on China, Brice. It seems as if that exposure for the industry has kind of settled in this 25%, 30% type of zone. Do you think that, that's sort of the run rate from here? Or you think that it will actually come down over time?

Well, when you say it like that, that makes total sense to me because I think China had an ambition to be self-sufficient from an equipment perspective and a semi cap perspective. And the China share of global semi device demand is around 30%. So if it settles in that ZIP code, that makes sense from a long-term perspective to us. Our most recent quarter, I think, was 24% from a semi systems and services business, so not counting display. So 24% and that's about the level we'll operate in the very short term.

And then as the leading-edge grows so quickly, it will probably be a little bit smaller, as we get out a few quarters, but that's the right ZIP code.

Got it. And then on just capital allocation, use of cash. It's interesting over the last decade, we have seen analog companies in semiconductors, right, get very high multiples because of their free cash flow generation and returns. Although I think for semi-cap companies is actually not as well appreciated, right? You have also done a lot of buybacks and had added a lot of cash. But how do you think about capital allocation? Do you think there are opportunities to do M&A accelerate your growth? Or do you think it's still return of cash that will be the primary use of the...

Yes, I definitely would not say return of cash is our primary objective. Our primary objective is to grow the business. When I think of it, Vivek, our business has a greater than 30% return on invested capital, I expect that to be growing and so I think all the finance people here will be thinking, well, if you have that kind of return on investment capital, it probably suggests that you have more projects that you could do at a good return, and that's certainly the way we feel.

We're building more lab space. We expect to be able to add more projects that will be profitable for the company. So the first priority is to invest in the business and grow the business. And then to your point, there will be significant cash flow to distribute to shareholders and no change in strategy there. We do that through our dividends, which we've grown nicely over the last 4 years and our buybacks.

Are there opportunities to do tuck-in sight improve? There's a lot of talk of adding more AI-like capability, right, in kind of different parts of the business? But are you seeing any of those opportunities come up?

Absolutely. I mentioned two acquisitions that we've done just recently. So I think we're active in that way. And those will be -- when we do those, that will be to add to the portfolio and add to our portfolio capabilities. Applied has a very great interconnected portfolio, and that's what we think helps us work out on the edge of the future technology with our customers, and we'll continue to build on that.

Great. On that positive note, thank you so much, Brice. Really appreciate your time.

All right. Thank you. Thanks very much for all those questions.

Thank you for coming, everyone. I'm Stacy Rasgon. Hopefully you know I'm Bernstein's Senior Analyst. I cover the U.S. semiconductor and semiconductor capital equipment space. And it's my honor to welcome our guests here today, Gary Dickerson, the President and CEO of Applied Materials.

Before we get started, I want to remind you if you have questions you'd like to ask during the presentation, you should have a link to our pigeon hole forum, where you can submit those, and we will have time for Q&A at the end.

So look, semi cap has been pretty remarkable. We've got a renaissance in the industry over the last several years, but it's really shifted into overdrive recently as AI has gone mainstream. The demand profiles that we're seeing are calling for -- I mean just simply more, more chips, more wafers, more tools.

In that world, like companies like AMAT, the contributions that they are bringing to that world are more important than ever as materials-driven innovation moves to the forefront of process technology development and enables the industry to satisfy this incredible amount of demand that at least for now, we still have. And to tell us all about what he thinks about it, it gives me great pleasure to welcome Gary and I've been looking forward to this conversational conference. So this should be fun. Gary, thank you so much for coming.

Thank you, Stacy. Really glad to be here. Always fun to hang out with you.

Yes. like -- maybe just to start on that point, I mentioned sort of a renaissance. So it wasn't that long ago that $100 billion in WFE was seen as just like aspiration. And we've blown through it, right? We might be 50% higher than that, like this year and growing potentially like for multiyears, I mean, you guys have talked about significantly increased visibility and a multiyear long-term growth cycle. Maybe just to start there, I mean, clearly, this is all driven by AI.

AI is sort of dragging everything in the space along with it, which certainly helps you guys. And -- what is AI actually done for wafer and equipment demand? Why are we seeing what we're seeing right now in this environment? And then maybe as you go through your -- think through your -- the business and the different segments and how AI is impacting all those, whether it's logic or DRAM or packaging?

Yes. I mean, again, AI is the biggest technology inflection of our lifetimes, really -- and we're using AI inside Applied Materials.

I want to ask you about that a little later, but...

Innovation, product development, operations, supply chain, our service, engineering, process engineering. And for me, the really big driver is top line growth. I mean, certainly, we're going to be more efficient and more productive. But I will bring multibillion-dollar products to market faster, even some later this year using AI technologies, and we will make our service engineers more productive so that we can deliver higher value services and increase the compound annual growth rate for that business.

And so I think that, again, this is just driving tremendous computing demand. And for Applied Materials, we're the leader in materials innovation in the most important and fastest-growing segments that are enabling AI. So leading-edge foundry logic, DRAM and advanced packaging. Those three segments are way more than 80% of the incremental WFE spending in '26. We think that profile continues going forward.

And you talked about all of the discussions with customers, I'm in very frequent discussions with customers on delivery times, and we can talk about that. We've done a lot to improve our operations and supply chain. So we have ramp agility. But the other thing, again, I'm meeting CEOs and R&D leaders all the time is the AI computing innovations. So one thing everybody is focused on is tokens per second per watt, tokens per second is performance. Per watt is power and cost.

And again, those three segments, leading-edge foundry, logic, DRAM and advanced packaging are the most important for chip architecture innovations. So again, Applied as the leader, where we've said we're growing our systems equipment business more than 30% this year like mid-50s gross margins and continuing to go higher. So again, that's making a major difference for Applied.

And then advanced packaging is the other segment that is growing very quickly. Again, those architecture inflections, how you connect computing components together and how you move the data is incredibly important for tokens per second per watt. So again, Applied has the most enabling portfolio. We're in deep engagements deeper and longer than ever before, because we are co-innovating with our customers. So we're out 10 years in the future, multiple technology nodes and Applied is the key enabler of those new chip and packaging AI architecture inflections with this materials innovation portfolio that we have. So I've never had more fun, Stacy.

We can see that. We can see that. I want to take those 3 segments one at a time. We talk about leading-edge logic and DRAM and packaging because I do think those are the -- I agree with you. I think those are the 3 at least at this point, are probably driving the majority of what we're seeing. And maybe let's take them one at a time, maybe to start with advanced logic and foundry. What is it that you're doing? You've talked a lot about things like gate-all-around and. What is it that AMAT is doing? And how is that actually driving the business today?

Yes. So again, for AI, every node in foundry/logic gives you something like 30% power, 15% performance.

Do they still do that, by the way?

Well, nearly that, I would say. And so if you look at the transistor, Applied is by far the leader and also for wiring, it's a multibillion-dollar business. And I tell everybody in your smartphone, you have a processor chip with 20 billion transistors, 60 miles of wiring, and it costs $30. So what we do in the semiconductor industry is we're delivering miracles every year. I mean that's basically what...

As I've said, I'm amazed any of it works at all .

And so again, for gate-all-around, which is how you process the data in the transistor, there are more innovations on next-generation gate-all-around -- in the future, they're going to stack [ NMP ] transistors for CFET. In the wiring, there are tremendous innovations. Customers are also putting the power lines on the backside of the wafer, which gives them better power performance, 30% area scaling. So all of those areas applied as those architectures are adopted, we'll capture over 50% of those available markets and gain share. And so what we've said for foundry/logic is with gate-all-around and backside power, it increases our revenue per wafer start about 30% -- so we're...

You gave one. I can't remember what it was.

Yes. So basically, the combination of transistor and wiring goes from $12 billion to $14 billion and then for 100,000 wafer starts, but there's also share gains for us. And one thing I would say that is also very different, and this is when you talk about DRAM or foundry/ logic that's changed in my time in the industry. I think in the past, Applied was more providing equipment to customers. But over the last 7 or 8 years, we've built a team of integration innovators that are as good as any of our customers.

And if you -- in a transistor, you have something like 500 steps and wiring as many steps. And so there's co-innovation and co-optimization across that flow and about 30% of our revenues are also integrated systems where we're combining multiple technologies under vacuum because some of these films are an atom or 2 atoms thick. And so if you go to air, you oxidize and you damage electrical properties. So I think one thing that is really important for Applied Materials in enabling these AI computing innovations is this portfolio we have, our ability to optimize, and that's why our customers are joining EPIC. That's why we have these very, very deep partnerships with these customers, enabling AI computing innovation. So foundry/logic is certainly an area where Applied has tremendous strength, and we will grow share as these new architectures are adopted.

Got it. I'm going to get back to the EPIC center stuff and .

Okay. .

But I want to talk about DRAM now -- you've picked up -- I can't remember something like 10 points of market share over 10 years in DRAM.

Overall DRAM spending. Overall DRAM spending.

And now, I mean, clearly, I mean, if I'm looking at like one of these AI racks and I look at like just the wafer area of the semiconductors within the search, I mean, HBM DRAM must be 80%-plus. There's a lot of DRAM that that's going to be.

Oh, no question. And then HBM, you have to start 3 or approaching 4x the number of wafers because you're stacking and you have some yield loss to have the equivalent number of bits versus standard DRAM. So I mean that -- obviously, DRAM demand is extremely high. And for Applied Materials, again, as you said, we gained 10 points in a little over a decade in overall DRAM share. And Applied is by far the leader in wafer fab equipment, process equipment for DRAM. And so there, again, for AI computing, people want really fast memory. And so in the CMOS logic, they're innovating with metal gate and new versions of metal gate where Applied, again, has very, very high share to go faster. They're moving to FinFET, where again.

So the logic portion of the DRAM.

Logic portion of the DRAM because, again, how you move the data is incredibly important. So those are all areas where Applied has tremendous strength. So -- and again, we're in these multi-node, many year engagements with these different customers. We have a great portfolio of products. And we're also in these high-velocity co-innovation relationships with customers.

So as they're innovating with the CMOS logic for higher speed, as they're innovating for vertical channel transistor 4F-squared or 3D DRAM. Again, we have very deep partnerships, just like we do in leading-edge foundry logic, where those innovators integration teams, architecture teams are working with Applied Materials creating those future architectures. So what we've said in DRAM is that, again, we're positioned as these new architectures are adopted, the CMOS logic, the 4F-squared, 3D DRAM for significant share gains, continuing what we've been able to deliver in the last little over a decade.

What kind of time frames are we talking about for like 4F-squared or 3D DRAM?

Well, I'd rather have the customers give those time lines. So people are extending 6F-squared. And then between the 3...

4F-squared will be a smaller cell size, so a higher density of the memory bits on the...

Higher performance, lower power, higher area density, all of those things. But again, the bets that people are placing are not exactly the same in terms of time lines -- so I'd rather have them talk about -- but that's coming over the next -- again, the CMOS logic is the most near term because high-speed memory is so important. And that's, again, a tailwind for Applied from a market share standpoint. And then most people are 4F-squared and then 3D DRAM in that kind of sequence. And that's going to happen over the next number of years.

And it really sort of brings -- you really are engaged, I mean, 10 years out.

Oh, absolutely. Yes, again, for these architectures to be adopted, this is where -- this is really the strength of Applied's portfolio. If you look at these inflections when we look at leading-edge foundry logic, when you go to 14A or 1.4 people use different terminology or you go to these DRAM architectures, they're somewhere between 5 and 10 key innovations that determine timing of those architectures.

Applied is by far the leader. If you look at 3D DRAM, really, the top 3 key technologies are all from Applied Materials.

So what are those technologies?

Well, I don't know if I want to go into exactly what all of those things are. But again, this is where we do this high-velocity co-innovation because in some cases, you need an order of magnitude improvement in the equipment to enable the performance, power, cost, yield of those new architectures. And that's where -- so for DRAM, again, high confidence that we're going to continue to gain share there. And that's from a memory standpoint, especially for AI, fastest-growing segment in memory. And again, Applied is in a great position. We can talk about EPIC again, how we work with people with all of these architectures.

We'll get to you next. But first, I want to talk about packaging because this is another huge -- I mean, look, as people -- Moore's Law slowing just which the transistor scaling is slowing and you have to do other things. Packaging is certainly one of the -- how do you put together these different chips in new and interesting ways. And for AI, clearly, it's enormously important. And you guys must be the biggest advanced vendor.

Yes, we are. So again, packaging, I think, is one of the most exciting inflections in the entire industry. It's really how you connect computing components together. And so I think Jensen and others, they talk about wanting to make the packages as big as they can to connect as many computing components as possible at the highest density because it's so important for tokens per second per watt.

So Applied, we have, again, a very unique broad portfolio. That business this year is growing more than 50% for Applied Materials, multiple billions of dollars. And again, we're also deeply engaged through the ecosystem on new packaging architectures. So if you look at an AI server today in 3 or 4 years, the way you connect those computing components are going to be very different than what you see today. So there, again, people want larger body sizes. They want higher I/O density. Applied has the leading portfolio in enabling those inflections. And we have deep relationships with companies that are currently leading and also companies that are bringing new architectures, new packaging architectures for AI to market in the coming years.

Is this like panel or?

Panel is definitely one of those architectures that people are focused on. Again, warpage is an issue for packaging. So if you want to have higher I/O density, dealing with that is very important. And then just getting a larger body size. I mean, if you look at...

Glass substrates and...

Glass substrates or if you look at some of our largest customers, tech symposium, they talk about a number of reticle fields that you can connect in a single package. And so this is really important technology. And again, Applied is the leader. We've been investing for a number of years. We knew these inflections were going to have an enormous impact on AI, energy-efficient computing. And so again, over 50% growth this year, and that will be a very high compound annual growth rate for Applied going forward.

And the visibility that you have now seems better than probably -- I don't know if ever.

Ever.

Is it ever? Yes. I can what you said are you getting like 8-quarter like rolling forecasts and...

Yes. Yes. So I'd tell you the visibility we have in the technology road maps because we're co-creating those road maps is better than ever. And also from a demand standpoint, again, frequent conversations with customers on delivery times. And so we do have 8-quarter forecasts from every one of our customers and longer commitments from the customers. What's really important for us, again, we did anticipate that WFE could grow a significant amount.

So we made investments to nearly double our operations capacity. But in the supply chain, you also need the lead time so that we have that ready to ramp. And so for customers, it's not only that visibility or orders sooner, it's also exact configurations because you need to know what components in the supply chain you need for customers. So for us, for us to be able to ramp to meet the need, again, we're getting much better visibility from our customers than ever.

Got it. And I guess that gives you -- people to talk about a multiyear.

Yes.

Sound pretty confident of that.

Yes. No, look, I think that what we've been seeing is that computing demand is continuing to increase. Recently, people are talking a lot about Agentic AI. And so Agentic AI is layering on top of the demand that was already there. And that's more weighted to CPUs. The CPU, GPU ratio is more weighted versus where we've been. And then also DRAM is up higher, NAND is up higher, what we're modeling is 90% of the incremental WFE is in the leading foundry/logic and DRAM. So again, and then after Agentic AI, you have physical AI, and so.

They never stop it. How much of a constraint is clean room space? I had this view that like as strong as this year is this is -- we might grow WFE, say 20% this year. I know you guys have not given a WFE number, but the growth is going to be strong. But it's a constrained year because we don't have the clean rooms and they're coming online.

I guess how many -- you guys always talked about you track that, how many fab projects are you tracking? What kind of clean room space is coming on. And I don't know if you have a point of view on where growth could be if like clean rooms right now were unconstrained. Like how much unconstrained demand is actually out there?

Yes. I don't -- again, we certainly have a forecast for where we think WFE is going, base case, bull case, bear case. And it's been hitting at the bull case, pretty consistently.

I bit objectively even the bear case probably isn't that.

No, absolutely. No, absolutely. So we're tracking over 100 fabs. As you can imagine, there are a lot of incremental spending in the areas that are the fastest growing. I said over 80% is in those 3 areas. So that's basically what we're seeing. And I think relative to -- obviously, with 8 quarters, we have great visibility into '27. And from my perspective, I think AI, again, really changes how we work, how we live. I look at -- I told you earlier how we're using AI inside Applied Materials. The ROIs are tremendous. And we can see already that we are accelerating product time to market. So I believe the demand -- this is a multiyear wave that Applied is in the best position because we're in the -- we're leading in the fastest-growing segments that are enabling the tokens per second per watt.

Got it. Got it. So let's talk about this co-innovation with customers. Tell me about the EPIC center. So it's coming into its own now.

It is. Yes, we've announced 10 partners for EPIC. We have TSMC, Samsung, Micron, Hynix, Broadcom. We have Advantest, SCREEN, some universities. You'll see more announcements coming. And again, this is what I said earlier relative to the chip and packaging AI architecture inflections is incredibly important for the ecosystem, but certainly for our customers because whoever is first to market with design wins, that has an enormous impact on their business. So Applied, we have these new innovations that will enable these new architectures in EPIC. And so for customers to have first access to those innovations, like I said, some of these technologies, you need an order of magnitude improvement or the architecture won't happen. So having their innovators locating.

How does it work? So they have personnel...

Absolutely.

Facility working with your personnel to developing their processes that they're bringing into production.

Yes. And again, it's really this co-innovation, high-velocity co-innovation concept. So we're -- we have customer reticles. We're focused on enabling those architectures, as you can imagine. And by the way, we focus on high velocity and secure innovation is incredibly important. So we firewall innovators between the different customers, and we've been doing this for a while. But now having those people co-located, especially, again, in Silicon Valley within 50 miles, I think, is 40% of the S&P 500. And so if you think about materials to systems innovation through the technology stack, which is more and more important when I talk about packaging or talk about chip innovations, those kinds of things, -- having these people co-located is incredibly important for velocity because you can move in parallel versus a serial process.

So our goal really is to accelerate the time to market significantly, incredibly important for our customers. And then for us, being designed in with our equipment and our advanced services for those architecture inflections, obviously, is really important for our growth going forward. And the other thing I would say that's a real advantage for us is our visibility into those road maps, where to invest if we have -- if we're creating those architectures, -- it gives us a great idea how to optimize our R&D. And again, for customers, not only is it important for their design wins, but their R&D efficiency. If you can bring these architectures to market because every one of them costs an enormous amount of money, there's an incredible efficiency. So this co-innovation, we have very deep partnerships through the ecosystem. And EPIC, again, this is where the key materials innovations for chips and packaging will first be available, and we don't have an infinite number of innovators. So having people co-located there really increases that velocity.

Stuff they're working on in EPIC, how far out is it?

Oh, gosh. So I mean, we -- again, we have incredible visibility. We know performance yield of every one of these different partners.

Stuff that's going to be in production like in 10 years or?

Yes. I mean we're out to -- I mean, N2 is ramping now. Next is A14. We're doing A10, A7, A5. So again, I talked about these miracles, Stacy, that we deliver in this industry. And so we're working on those miracles that enable those future architectures or again, the DRAM road map. I mean some of these technologies are extremely compelling for the AI computing, but they're also incredibly challenging to create the innovations that enable those architectures.

So we're working 10 years out in the future, multiple technology nodes. And again, for us, it's being designed in with our equipment and services. Also, it creates a deep strategic relationship with these customers. So when I'm with the CEOs of our customers, we're focused on those future technology nodes. And for them, it's design wins in certain time frames, technologies that they need to hit those road maps that is really essential for their business.

Do you need other players' tools in there as well? I mean, I assume there's litho and other things in.

So we do have partners that are -- some announced, some that aren't announced relative to those innovation flows. The great thing for Applied is we have more of those innovation technologies than anybody -- it's not just the strength of our -- people talk about deposition. Again, we have technologies that are completely unique, epi, PVD, in addition to the rest of our deposition portfolio. So creating the materials. People really also don't understand the modification of the materials. So that's a multibillion dollar business for us and then there's shaping and then also the -- our -- we're the world leader in eBeam technologies for analysis of the materials.

But that combination of technologies enables you to create these future transistors or wiring or memory or packaging, all of those things. So I think this portfolio strength that we have -- and the way we're engaging with customers is really unique versus where we were, let's say, more than 5 years ago.

Yes. Now that makes a lot of sense. I want to talk about the services business. And I guess maybe you can talk a little bit about how it's changed over the last 5 to 10 years. I know there's a lot more you talked about recurring. You talked about use of AI. I mean, this has got to be an area where we're -- so how has this business changed? And -- maybe talk a little bit of what the services brings to you and to your customers? .

Yes. So services, right now, we're growing high teens compound annual growth rate. It's a little over 20% of Applied Materials revenue. And we said through cycle longer term, we'll grow mid-teens. I think there's an opportunity for even to drive that faster over the longer term. As you can imagine, for customers, one thing that's super important right now is output per square meter. Again, so...

Square meter of fab space.

Absolutely. The more output they can have, the better -- so yield is really important. Edge die yield is really important. And Applied, I really love this part of our business. We've driven tremendous service innovations. Over 2/3 of our business are service agreements and prescription or subscription revenue. So that's been growing actually faster than our overall service business. And we have over 35,000 chambers connected to AI servers in customer fabs and the majority of those are remotely connected. So that is really important because, again, as we're bringing these new service technologies using these digital solutions, we're optimizing output.

Chamber matching. That's a really important technology. So there are hundreds of thousands of opportunities per year. When -- and again, we're using AI to optimize with all of these different parameters for all of our process tools. But then once we make one golden in high-volume manufacturing, you want to make them all golden. You want the biggest process window to give you the highest yield.

And so we're innovating there with these service technologies. We're also innovating with predictive models versus reactive corrective maintenance. Again, all of those things, I think, are great opportunities for us to grow this. What we said is, again, kind of mid-teens growth rate going forward. It's a little over 20% of Applied revenue today. And I really believe, again, the value of the services are increasing as you're ramping these complex new architectures, great opportunity for Applied.

Got it. How much of that growth is like -- because it's an installed base business with just the installed base growing. And some of it is this content per tool, I guess.

Revenue. Yes. Yes.

How do those contribute to this? And are either accelerating? I got to imagine maybe the revenue per tool.

Revenue per tool is growing faster than the installed base. And again, these services as we're -- and again, this is where I'm also focused on AI. In some cases, where the dollar per task or hours per task were improving many, many, many times. So for me, the real important part is not are we saving thousands of service engineers. The important part is I can deliver high-value services and scale that much faster and drive the top line growth with those technologies. So again, really fantastic opportunity for Applied.

Got it. Got it. I want to talk about China. There's a few things going on there, clearly. Number one is there's been regulatory and other issues that have impacted your abilities as well as your peers to sell everything you would like to sell in.

Yes.

Which is -- there's also the growth of local competition. I got to be honest, like one of the biggest questions I get is what is the impact of some of these local players. I think etch and deposition are 2 of the areas where there's probably like a larger piece of the TAM is satisfied by the local guys. And as I would say, these are real companies there. They're not jokers, right? Just maybe talk a little bit about what is going on in China. What do you see on the competition front there? I guess anything you can tell us on the regulatory front one way or the other? I think there's some other stuff coming in the MATCH Act and other things that may be coming. So how should we think about China?

Yes. So '25, there was a large increase in restrictions. We had said that our business was restricted at about 10%, more than double in '25. So that did have a big impact.

China sales were still very good.

Yes. And so for today, for us, it's kind of mid-20s as our semi equipment and service business. So for us, China is really an ICAPS business, the IoT, communication, auto, power sensors. And this is where Edge AI is also a driver for that particular market. So in China this year, and in ICAPS globally, we believe that business will be flat to slightly up. So...

How much of your foundry logic is ICAPS today, by the way?

Well, we haven't -- we have said in the past, it was kind of equivalent. But actually, again, going forward, the leading edge driven by AI is growing much faster. So we think over time that ICAPS will grow kind of mid- to high single digits. And if you look at actually data center, wafer starts, this is an interesting statistic. Our -- will pass smartphone wafer starts, we believe, next year. So these large die, you're driving a tremendous growth.

So again, the growth rate for the leading edge is faster than the ICAPS market. But again, for us, that business is kind of flat to slightly up. We do have innovations. We started our ICAPS group. I remember April 12, 2019, because I thought that was a really good strategy for Applied. We have a great portfolio there. We're bringing new ICAPS innovations to market, new products that expand our total available market, new products in large segments that improve our competitiveness.

So we are driving innovations there. We have deep connectivity with customers on architecture innovations in power electronics, this co-innovation, sensors, photonics, a number of different areas where we're engaged. So I think, again, that business I have high confidence that we'll perform well there over time, but the growth rate is not going to be as fast as the leading edge, which is driven by AI.

Sure. To go back to China, the competition front, the low...

Yes. So again, that's where we're bringing these new products to market. And -- if you look at where we were able to compete, if you take out the restrictions, we performed well in China. And I think that with the pipeline of products that some later this year, some over the next 24 months, we'll expand our total available market. We have products that will improve our competitiveness. So I think where we can compete, I feel good about our ability to compete.

Can't compete everywhere, unfortunately. .

Unfortunately, yes. .

I want to talk about margins. So -- and I said this to Tim yesterday, but I always ask you the same question. It's always why do your gross margins have to start with the 4. And I don't have to don't anymore, don't start with the 4 anymore. I guess what's gotten you there, number one? And then number two, how should we think about the evolution of that margin structure going forward? There's a whole bunch of drivers. There's clearly operational efficiencies that you guys are driving. There's mix. Maybe there's pricing, like I don't know like for sure. So like how do we think about that?

Yes. I think that if I look at our conversations with customers, they're very different. Now it's about delivery times, access to our technology innovations to enable the new architectures, -- so we have been growing margins pretty consistently. I think you're going to continue to see us growing margins, both gross margin and operating margin a fair amount as we keep going forward. And so again, it's really the value that we're delivering in pricing. We have been increasing pricing. I think one thing that...

You've been increasing pricing of like stuff you already sell? Or is it more on like new.

Both. Yes. And so -- and again, different for different situations in terms of how we optimize all of that. But again, what's really important for customers is that we are able to make the investments in our operations, supply chain and infrastructure for the deliveries. And what's incredibly strategically important for them is the investments in the enabling technologies for these AI compute architectures for chips and packaging. So again, those are the discussions that we have. And then for me, what I think about is winning these architecture inflections and creating value, if we're creating value, we also have a great opportunity to capture value.

And one other thing I would say in the ecosystem, the profit pools are way larger than they have ever been. And so again, that gives us an opportunity to create value and share value with our customers. And again, I feel really good about our positions in these future AI chip and packaging architectures -- and we have an incredible pipeline of numericles that will enable these architectures and give us a great opportunity to create and capture value.

Got it. Maybe on the value capture. You talked some about like integrated tool sets. So I think you said some of it. So does that percentage go up over time? I have to imagine is this stuff features get smaller and this stuff gets harder. You're not really just selling a single process. You're literally selling a solution. We want to create a film that's this thick and it's conformal, and this and it takes 10 different processes to -- and then how do you -- like how do you think about capturing that incremental value versus sharing it with your customers? .

Yes. So I talked about the 60 miles of wiring, which is kind of mind-boggling. When you think about moving data with almost no resistance, through a wire, and it's even going higher than that. So some of these platforms, we have 7 technologies integrated under vacuum because, again, -- if you go to air, you can oxidize those few ADAM interfaces. So that business of our business is growing.

And the the number of technologies that we're integrating is also growing. We announced some new ALD technologies. And then, again, in those technologies, the surface interface is really, really important for electrical properties. Our PVD business, combined with selective ALD and CVD and copper reflow and all of these kinds of things.

That's a unique capability Applied has to enable these future architecture inflections -- and then it's not only the integrated products, the strength of our portfolio is our ability to optimize across these modules. So there's co-optimization to enable this tokens per second per watt across the portfolio. But the integrated is about 30%. The number of technologies we're integrating in the platform is increasing and super strategically important for customers for these architecture innovations.

Got it. I guess on that note, like I always think about the semi cap industry, I mean it's competitive that everybody kind of has their own niche. But at the same time, you clearly do compete -- I'm going to leave the Chinese guys out of it, but -- you got the Lams and the Tokyo elections of the world -- what is it that customers are looking for when they're choosing a new solution? Like what is that -- do -- what does that bake-off process look like? When does it happen? Is it the new node transition what is it that customers are looking for when they decide to actually award the process of record to you versus your competitors? .

Yes. So I think the most important thing for customers is the race for these new architectures. That is the most important thing. And whoever -- so when they're competing for design wins, they care about performance, yield and reliability. That is the most important thing. And when their customers are evaluating them for design wins, if you don't have performance yield and reliability, they don't care about the cost because they can't risk their system architectures on those technologies.

So for Applied, we're working through the whole tech stack and working with these customers on these new architectures, that is the most strategically important. I mentioned high-speed DRAM with the CMOS logic or any of these different types of inflections. So the first thing for Applied is again, connecting with them on what's critical for their design wins -- and then inside of that, you have a portfolio of solutions where Applied is driving innovations in each one of those different pieces or those integrated platforms.

But that connectivity, I think, is more and more and more important. So when we're engaging with someone on an N plus 1 architecture or an N plus 2 architecture, we're really focused on the electrical, like resistance, capacitance, threshold voltage or high-speed memory or those kinds of things. And then the technologies that enable those key performance, the tokens per second per watt and then there's a portfolio of Applied solutions in key unique Applied solutions that enable those technologies.

So again, the great thing for us, we have leadership -- we're in great positions in the fastest-growing segments that are enabling AI computing in leading-edge foundry logic, DRAM and packaging. So again -- and then the other part of our portfolio that is actually growing very fast -- this is one of the fastest growing businesses in Applied is our inspection and measurement business.

So we said our semi business is growing over 30% this year Actually, that part of Applied Materials is growing much faster. We're the world leader in eBeam technologies. We have the lead in resolution and imaging speed with cold field emission. That will be over $1 billion in revenue this year. And we have an incredible pipeline Stacy of new products that are coming over the next couple of years. So I have high confidence. That one is one of our fastest-growing businesses. The revenue growth there is going to be multiple billions of dollars. But the other thing that's really important is that when you're -- when you're building all of these new architectures, the learning rate is crucial.

So Applied's eBeam technologies, we can see into those architectures and collect data, orders of magnitude faster. So again, that connectivity with our leadership and the materials innovation is also strategically important for Applied when we're engaging with these customers on the architecture innovations.

That makes a lot of sense. -- one other segment that you don't talk that much, it's small, but like why do you still have a display segment. What strategic value does that play for you? .

Yes. So display today is a little over 20% operating profit. We -- and it is growing at a decent rate. The biggest opportunity there for us is we do have an opportunity to enable a new OLED architecture for IT and TV. And so there has been a technical barrier. So today, you have single-digit adoption of OLED in those segments of the market. Applied has come up with an innovation. We're working with the leaders that overcomes that limitation. So it would take some time for that adoption. But I believe that, that can be a good growth rate for Applied, expanding the operating profit.

And so that's one thing. Just that business by itself, there is synergy when you talked earlier about panels being processing. So if we think about our leadership in handling glass, or all of the glass processing applied is the global leader in those technologies.

We're taking that learning into these new packaging architectures. So there's synergy -- we have -- by the way, this is an incredible team of innovators inside Applied Materials. So there's that synergy with where packaging innovation is going in the future.

We've got a few minutes left as you go to the lightning round .

Sure. Let's go .

And if anybody has other questions that you want to submit, you can do that. Are there any areas where you think you're participating less than your fair share in the super cycle due to competition or anything else?

No. Again, I mentioned earlier, we have really great visibility in the architectures, chip and packaging architectures that will enable AI computing innovation, the tokens per second per watt. So I feel great. And again, the really great thing for Applied is that -- we have incredible visibility because we are working and co-innovating and co-creating these architectures through the tech stack and through the ecosystem.

So we are investing in the areas that we think are going to drive the highest growth. We've made some small acquisitions. It's not possible to do the large acquisitions with the current situation. But there are areas where that can add hundreds of millions of dollars of growth and create synergies with the rest of Applied Materials. So I feel really good about that.

Got it. Here's an interesting one. Is Huawei's logic folding economically real or road map theater.

Well, I don't know if I want to comment on that. What I would say is that there's this concept of design technology co-optimization. Some of our largest customers have talked about that. You're moving the power lines from the front of the wafer to the back of the wafer. You're getting a 30% area savings, better power and performance. You have isolation walls in the transistor -- so there are those kinds of innovations that are being driven in the industry.

And so I would say this is kind of along the same lines, and one of our largest customers, they talk about how the percentage of their innovation enabled by design technology co-optimization, going up significantly, which is, by the way, great for Applied Materials. But that particular one I think it's too early to tell. I mean you have thermal issues, yield issues, many different kinds of things that are yet to be determined.

By the way, that's using bonding technology. Applied has the leading hybrid bonding technology and an integrated platform. So as those architectures are adopted, we also have innovations there.

I think it's fair to say that those particular architectures are not unique to Huawei or the industry is looking at.

No, if you look at even like future transistor technologies, you're stacking NNP transistors. So it's really going more vertical and -- or like, say, power lines or transistors or all of those kinds of things or CMOS bonded to the array in memory. So separating out those IP blocks and connecting them in different ways is a trend that the overall industry is focused on.

Got it. Maybe one more. Do you think there's a disconnect between spending expectations from the hyperscalers and the labs versus what the semi industry can realistically deliver.

I feel very good about demand going forward, I would say. I do think, again, this is going to change how we work, how we live AI is going to be an enormous inflection. That's going to drive tremendous computing demand and the need for computing innovation and tokens per second per watt.

And Applied, like I said, we have made investments in our operations so we can scale our operations. It's really making sure that we have that alignment with customers. We have to have that early visibility to ramp our supply chain. So I feel good. I mean, we've made enormous improvements in operations and supply chain over the last few years, and we're in a way better position than we've ever been in that aspect.

Got it. So we've got about 1 minute left. I'll close out the way I always do. We've got a full room here, Gary. Why should they buy your stock?

We're in a great position. The fastest-growing areas in the industry, leading-edge foundry, logic, DRAM, packaging, that's over 80% of the growth in WFE. Those are areas where Applied has clear leadership. And those are the areas that are continuing to grow significantly over the next several years. And as I mentioned, our connectivity, our relationships with our customers are deeper than ever and critical to them winning designs with their customers.

So those future AI chip architectures, packaging architectures, Applied is in a great position to gain share, create value and capture value. So I have high confidence the growth rate for Applied is going to be very strong, both top line and bottom line going forward.

Got it. I think that's a wonderful place to leave. Gary, thank you so much. .

Thanks, Stacy. Thank you, guys.

All right. Good morning, everyone, and welcome to the third day of JPMorgan's 54th Annual Technology, Media and Communications Conference. My name is Harlan Sur. I'm the semiconductor and semiconductor capital equipment analyst for the firm. Very pleased to have Tim Deane, Senior Vice President at Applied Materials. Tim heads up Applied's Global Services business, what we call AGS, represents about 20% of Applied's overall revenues, drives a stable, but very high-growth services franchise, which is a key part of the systems-level solutions they develop and deploy to the world's leading semiconductor manufacturers.

Tim has been Applied -- been with Applied for over 30 years, in charge of AGS for the past 3 years and previously worked in the Semiconductor Systems Group for more than 25 years, leading the field operations team, business management, strategy and customer business and planning team. So Tim, thank you very much for joining us this morning.

Yes. Nice to be here, Harlan. Thanks.

We're going to start off at a high level. The team just reported very, very strong results and growth outlook last week. So I'm going to start off there, but we are going to touch upon Tim's AGS franchise because it's just -- it's a great ratable, highly profitable cash flow generative franchise. But starting off at a high level, team reported strong results outlook last week.

On the better revenue growth outlook for this calendar year, I think you guys are now calling for 30% growth in your systems business versus your prior growth outlook of 20%. On the incremental upside this year, is it being driven by new brick-and-mortar greenfield programs being pulled forward? Or are customers just accelerating technology migrations on existing capacity? Or just solidifying plans and now sort of locking in sort of tool slots, right?

And so trying to figure out like what drove that sort of incremental upside over the last 90 days? And then can you just talk about which end markets, advanced foundry logic, DRAM, advanced packaging contributed most to the sort of incremental improvement in the outlook?

Sure, Harlan. Happy to talk about that. So there's -- I would say there's not fundamentally a pull-in from '27 into '26. There's a lot of brick-and-mortar being built out there, greenfield facilities. So there's a pipeline of projects that are coming online. To the extent that customers can accelerate them, they obviously want to do it as quickly as possible given the environment. But there's a lot of creativity going on in the existing spaces out in the factories globally today.

There's at least one case where a customer procured a completed shell from another customer. And obviously, they're going to put product into that as quickly as they can, which helps us with our business. But the other areas that we see are primarily in-fab changes. So if customers are looking at their full production lines, if they see pinch points in their production, maybe an older generation tool that's a lower throughput that's limiting the output of the factory, we're seeing lots of cases where customers are coming to us and saying, okay, we're going to remove this platform. We're going to replace that with a higher productivity platform, which is good for wafer fab equipment.

Additionally, we see cases where -- or there's at least a case where a 200-millimeter factory was being shut down and repurposed into a packaging factory. So you were asking kind of which areas do we see the growth. Packaging has got intense growth pressure in demand in end markets. So we're seeing customers being as creative again as possible to get capacity in as quickly as possible.

And then we see another example where there's an older generation 2D NAND factory that's being converted to a DRAM factory. Again, DRAM, advanced packaging are kind of themes that you will hear over and over again where customers are trying to get more good die out. So that creativity is leading to these incremental upsides that we see. And then just from a services perspective, interject, as we work with customers across all their facilities, I think in every place that touches AI, those customers are coming to us saying, how can we optimize output for the factory, which is kind of driving the service business as well.

Yes. I think that's a little bit under the radar screen for most investors is that we all tend to think about greenfield capacity. We all tend to think about technology migration. But as you pointed out, there's a ton of installed capacity out there where throughput may not be as optimized, yield may not be as optimized, right? And any little improvements in current factory throughput or defect density improvements can have a pretty significant impact in terms of output and good die per wafer, which ends up being millions and millions of dollars for their customers, right?

No, that's a good point. If a fab is fully installed and that's operating, every percent of utilization improvement that you can make in a factory is a huge lever for productivity for customers. And as we partner with them, particularly in the factories, whether it's with upgrades to existing tools, placements of key tools or just service optimization, these are all kind of win-win scenarios that we see with customers. So it's a great environment out there to be working together really closely.

And at earnings, the team called out having historically high visibility, 8-plus quarters type of visibility to your backlog, customer forecast. And this gave the team the confidence last week to qualitatively talk about continued strong growth profile in WFE spend '27 and '28, right?

And in line with that, Brice talked about the Applied team having 100-plus projects, programs that are on your radar screen. These projects sit at different stages, some near-term ramps, some longer-term commitments. Some of these programs are technology migration focused, greenfield build-outs. What's the rough mix of this 100-plus project pipeline, right? Is it technology migration more biased? Is it greenfield capacity biased? And how does this pipeline sort of convert over the next, call it, 2 to 3 years?

Yes. So just the first part of your question about the 8 quarters visibility, I think -- those of us that were around post-COVID during that ramp, there was an awful lot of learning that happened during that period. In my prior role, I was kind of at the nexus of that with Applied working with customers as they were trying to kind of plan capacity on what was kind of an unexpected ramp.

This time, it's a little more of an expected ramp because we know with AI and the Agentic components of AI now that are coming online, there's just huge demand across, in particular, leading logic, DRAM, HBM and advanced packaging. Those are the ones that we see as kind of the primary drivers. So it's in our best interest as a supplier and a customer to collaborate as closely as possible because we want to make these transitions smooth.

Bringing up these factories is extremely complicated. We or me, I'm hiring people across the globe in order to do installs, meeting customer schedules. So the collaboration is coming naturally, I would say, at the moment, particularly with our largest customers because it's in our mutual best interest to do that.

Now on to the pipeline. So when we look at the pipeline, 100 projects is a lot. And I think Brice also mentioned, 10 or more that's come fairly recently. The biggest part of the greenfield capacity that's coming online is actually in the advanced logic and DRAM spaces because there's just not enough physical capacity today. There's so much silicon that's needed for these advanced technologies that it's really being driven by the greenfield space.

I would say advanced packaging is in the same boat, so to speak, just because we're moving from something that's been relatively straightforward for decades, we're moving kind of the front-end type processes into the back end, which is, again, requiring more physical space in the factories.

NAND is a little bit different. You get so much leverage on a NAND wafer whenever you upgrade. So as we're present in all the NAND factories around the world. We see the wafer starts have kind of dropped a little bit over the last couple of years in NAND as customers have converted from one technology to the next. So it's a little more complicated. The technology -- the wafer starts drop a little bit, but the bits go up significantly.

So there's a lot more leverage that you get with upgrade NAND as compared to the other spaces. We've done some analysis, a typical, say, HBM wafer versus a NAND wafer, you're getting somewhere between 40 and 50x the bits on kind of apples-to-apples comparison for an SLC NAND. If you go to quad level, you're quadrupling that. So we think there's just a huge amount of capacity, and it's very efficient for customers to do that as conversion. Now I think there will be some greenfield NAND over time, which is great for us. We're very -- we're actually kind of excited about that as well. But we just see the leverage in the other 3 spots for greenfield.

Got it. And semiconductor industry is on track to drive over $1 trillion in revenues this year. I mean it was just a couple of years ago that we were all talking about a $1 trillion market by 2030. And on your recent -- these are very valuable sessions, but in your recent Master Class, you brought up an important insight, which that the mix of leading edge -- the mix of spending was longer term thought to be 1/3 leading-edge foundry/logic, 1/3 memory, evenly split between DRAM and NAND and 1/3 sort of mature and specialty analog power products, which you guys call your ICAPS business.

But given the rapid growth AI and data center compute, networking and memory, which biases strongly towards leading-edge logic and memory architectures, there's now a strong divergence, it feels like in growth between these various segments. So help us understand like how do you see -- is there a new way to kind of think about the mix of spending by markets, not only this year, but over the next kind of few years?

Yes. So for this year, we think about 80% of the incremental wafer fab equipment spending is really in those 3 areas that you talked about, the advanced foundry logic, DRAM and advanced packaging. And that really is driven by the AI demand. If you look at the various waves that have come up over time with compute and then to mobile and now into AI, there are different focuses for each of those areas.

ICAPS is still a very important market for us. And I would say we're -- I think Brice made the comment that we're incrementally positive on it compared to where we were, say, 3 to 6 months ago. We think it's kind of flat to up both China and non-China. But the bulk of the spend is really going to be in those 3 focused areas where Applied is #1 in advanced foundry logic. We're #1 in kind of DRAM in a lot of areas and then also for the advanced packaging.

Again, going back to the -- you guys held a great Master Class on transistor and interconnect architectures, right, on current and next-generation platforms. One of the big differentiators in my view, at the transistor level is Applied's what you guys call integrated material systems, right?

That's right, IMS.

So various aspects of the gate formation or the interconnect formation, you've got a tool that integrates up to like 7 different processes into a single platform, right, surface treatment, etch, deposition, clean, anneal, integrated metrology and so on, right? And these integrated solutions now represent about 30% of Applied's overall system sales. Obviously, the benefits of IMS is not only technology enablement, but as you could appreciate, better defectivity, better -- higher throughput, lower overall system footprint, it seems like a big differentiator for Applied, but why is it that the mix of IMS-based platforms hasn't increased that much over the past several years? And how does IMS create opportunities for your AGS franchise?

Sure. So I do think IMS is one of the things that's really -- it's unique for Applied Materials, right? We have a very broad connected portfolio of products. And the ability to bring them together into a single mainframe where you don't break back and you don't expose the wafer to ambient is, in some cases, it's very critical. I think in that Master Class, if I counted correctly, there were 4 IMS platforms that we talked about, right?

There was Trillium that you mentioned for gate formation. There was an Epi and etch combination for forming the contact that was talked about and then 2 versions of copper berries integrated tools for the super tight pack lower layers is one configuration and then kind of the mid-layers is another configuration.

So from Applied's perspective, we will work with customers as closely as possible to integrate steps that make sense to integrate because it does drive complexity on a single platform, as you can imagine. What used to be a single PVD platform Endura now is a 7 different applications. So managing that platform in production and just kind of redundancy on the platform, you want to optimize for where that's really needed. And it's really these absolutely critical steps that we do that.

The other thing we do at Applied, again, because of the portfolio is we co-optimize applications. And that's another big piece of the business. So they can be sequential steps or nonsequential steps that still need to be co-optimized because they interact across multiple layers. I think a good example of this is we're seeing great synergy between our PDC group, our metrology inspection group and our development teams, in particular with PROVision, where you can do thousands of measurements across the wafer, we can optimize processes faster than we have in the past.

So we tend to look at it pretty holistically, where it makes sense to do an integrated platform with a customer, if that's the demand, then we will do it. And we are constantly co-optimizing applications kind of into the flow where it makes sense to give the customer a little more flexibility. So it's not really about increasing that number. It's really finding the optimal space in the flow. But again, I think this is one of the areas where Applied is truly unique in the industry because of all these capabilities that we have across different product lines.

And complexity -- system complexity in my mind, means good things for the services organization that supports that. Is that a fair -- is the service intensity a bit higher on these IMS platforms?

The service intensity is higher on the IMS platforms for the reasons that you stated. If you think about -- just think about parts, if you had a 7 or 8 chamber, 9-chamber Endura that was all PVD, there's a lot of commonality in the parts. If now you have an ALD and an anneal and you got CVD and PVD, those are all kind of unique capabilities. So it causes us to prepare differently with customers to support those tools to make sure they're getting maximum output in the factory.

Your DRAM revenues were $1.7 billion in the most recent quarter, up 18% year-over-year. The team has roughly gained 10 percentage points of WFE-related DRAM share over the last decade. The next 2- to 3-year DRAM road map is meaningful for Applied, right?

Yes.

6F-squared continues with its high-volume ramp on the next-generation 4F-squared and then eventual move to 3D DRAM architecture, Applied has been pretty consistent that materials intensity increases meaningfully at 4F-squared and especially at 3D DRAM. Can you just frame for us the share opportunity for Applied in 4F-squared specifically versus where you sit today at 6F? And is the right way to think about it, given the historical share gain track record that maybe Applied gains another, call it, 3 to 5 percentage points of WFE DRAM share at these upcoming inflection points?

Yes. So let me start by putting a little plug in for Master Class because on June 25, we are going to do a DRAM advanced packaging masterclass, kind of the level of detail that you saw with the advanced foundry logic. With 6F-squared, in particular, we see there's a lot of runway. Customers are pushing those technologies. Again, the demand -- Manish is coming up next. I'm sure he's going to talk about the demand in the end markets is pretty incredible.

So when you have a technology, you want to extend it as long as you can. And the place where we see the road map going for 6F-squared is really right into the sweet spot for Applied's portfolio. There's going to be more advanced wiring. We also see that the transistor formation is moving more to the advanced style that kind of moved into advanced foundry logic a few years ago. So we're able to take that knowledge, the experience on the tools, the productivity and cost down benefits that we've seen by optimizing those tools and move those into the DRAM into the next couple of years.

When we start talking about 4F-squared and 3D DRAM, again, we just -- we see that the materials engineering intensity on those applications going forward is also moving into the Applied space. And this is one of the fundamental reasons why we created this EPIC platform. So we've announced a number of customers, Samsung, Micron, Hynix, TSMC last week. Advantest is joining as a partner. We have universities coming in, RPI, ASU and Stanford. And just this morning, we announced that Broadcom is joining the EPIC platform as well. They're specifically advanced packaging -- they're specifically interested in advanced packaging.

And these processes today are so complicated, it takes an ecosystem to essentially bring these to market. So I think there's recognition that partnering in this vehicle, which is EPIC for Applied is going to be a way to accelerate that. And our goal is to take what was a 10-year development pipeline and ultimately cut it in half by bringing everybody into the same facility and co-optimizing in an environment where the innovator is there, the customers are there, the customers' customers are there, industry partners are there, just to fundamentally change the way that development is done.

And from a service perspective, I view this as the best thing ever because historically, when we've done development with customers, we've done a demo in our facilities, a couple of wafers or chiplets. Now we're going to be developing a process with a customer in a facility where we can put our AIx capability around it, our eco solutions around it. We'll have our service folks there working on developing a fingerprint on these brand-new tools so that when a customer puts it into their R&D facility, we can fingerprint it immediately, bring it up faster, increase speed to market for customers.

So it's a whole ecosystem from the upfront R&D process pure research, all the way through how do we optimize it and get it up into production faster for customers in their R&D and ultimately in the HBM facilities.

Yes. No, I think now is actually a great time. So let's focus on your AGS, your services franchise. As the Head of Applied Services business, you're in contact with all of your major customers. I mean what are their priorities right now? And how is the Applied team helping them out?

So the priority -- in a word, priorities are output. It's good die out. And so it's not just optimizing throughput on an individual tool. It's literally optimizing good die out on every single wafer. So we're leveraging some of the capabilities that we've been developing over the last 10 years in AGS, specifically the AIx platform, which is actionable insight accelerator. This is the way that we bring AI, predictive models, digital twins, recipe optimizers.

This is the way that we bring these to customers in the factory on their tools. So we're working very closely with them on a project basis and a contract basis to optimize output and again, primarily to optimize good die out in this environment. So it's great for us given 2/3 of the AGS revenue today is driven by long-term contracts. So we've got really tight relationships with customers, and they're leveraging the footprint that we have in the fab to help them drive output.

2/3 of the business, long-term contracts and close to high 90s type renewal rates.

So we publish at a greater than 90% renewal rate when our portfolio and services is pretty broad. So we have kind of basic level services, warranty-like services. And then we have managed and performance services, which we really co-optimize with customers, how they manage individual tools and we try to drive output. When we look at managed and performance, the renewal rates are actually quite a bit higher even on those levels. So the stickiness of this because of the value that it provides customers is quite high. And this is a change that we pivoted to about, I'd say, 10, 12 years ago, shifting to more of an outcome-based services type solution. And yes, it's growing quite rapidly.

Your services franchise, 20%, 22% of the company's total revenue base, very, very strong recurring revenues, as you just articulated, within striking distance of 30% operating margins. Through the first half of your fiscal year, AGS is driving 17% year-over-year growth. You did raise your long-term growth outlook from low double digits sort of percentage CAGR to sustainable annual growth in the mid-teens range, right, and potentially higher at least this first year. What's driving the higher growth outlook?

So if we look at the business in the sort of the component parts, so the contract business continues to grow, and it's growing faster in the managed and performance space, which is great. That's exactly what we want with customers because we think that's where we can provide the most value. Those contracts, we are -- I think our average contract length at the moment is about 2.9 years. And the sweet spot for us is somewhere in the 2.5 to 3-year range.

So we're kind of right where we want to be from a visibility standpoint. It helps us plan better for parts and people and performance with customers when we're in that range. So that piece of the business, we're very, very happy with. The next component, I would say, is sort of our on-demand business, and that's on-demand spare parts or if people need contracted labor to come in and do some work. That's a little more tied to factory utilization.

The factory utilization is high, we see the on-demand parts going up. So that's driven it up a little bit more in the near term in combination with this continued strength that we see in the services business. And then the third component of the business in AGS, which we don't talk about too much is we have a software franchise in AGS. So the manufacturing execution system, Smart Factory, the SIM system, Applied is actually the largest supplier of factory-level software for customers.

And that's a space that we see a lot of opportunity going forward, particularly as we think about what's happening in the larger market with AI and Agentic AI. So that platform is an opportunity for us to engage with customers in different ways than we have in the past. So we actually see strength in all 3 of those, and that's why we think there's an opportunity. I think Gary mentioned in the near term to be above that mid-double digit for the year.

Your installed base of tools last 5 to 10 years has been growing at about a 6% to 7% CAGR. Your services business has been growing up until recently, 11% to 13% CAGR. Now you're forecasting mid-teens as we just talked about going forward. Installed base is accelerating. The team is also seeing good traction driving attach to more higher value-added solutions. You talked about things like AIx. I'm wondering how much of this is also contributing to the AGS sort of growth outlook, right? Last, we got an update from the AGS team. Recurring revenue per tool, which is sort of reflective of capturing more value-added capabilities, recurring revenue per tool was growing at about a 5% CAGR. Has this value capture rate actually gone up due to capabilities like AIx, remote diagnostics, digital twins and things like that?

Yes. I think the way we look at it is we are -- we have an innovation pipeline that we think of in AGS. And this consists of these AIx capabilities, supporting these IMS tools, which are different for customers than an individual tool. And then there are other capabilities, cobots and other things that we kind of have in the portfolio that we can deploy to support customers. So as we aggregate these capabilities, what we see is customers are moving up the value chain with respect to the types of service that they're buying from Applied. So more into managed, more into performance type service because they're getting more value out of it.

The other thing I would say that's really important to consider is the process windows today that our customers are dealing with on these really advanced nodes, whether it's memory or advanced logic, they're really small. I mean they're very tight. So part of our focus the last couple of years has been to find ways that whenever there is an event that happens in the factory, first of all, we want to take corrective maintenance events, extend them out. We want to move corrective to preventive.

And we do that with these predictive models that have been created because the ability to take a corrective maintenance activity, move it into a predictive means that now we can work with a customer to schedule that. So they're with moving through the factory, we can look at where there's an optimal time to take the tool down and bring it back up. But it's not just basic maintenance anymore because the process windows are so tight. So we created a capability that we call CROSSMATCH a number of years ago, and we've automated it in partnership with our business units, and it allows us to do a quick fingerprint of the tool before it goes down for maintenance.

When it's a predictive model, we know exactly what we need to do. So we've got the folks trained to come in and do that change. We bring it back up. We refingerprint it and re-center the process so that you get the maximum amount of time from maintenance event to a maintenance event. And this is incredibly important for customers because it allows them to run the tools a little bit longer. And in this environment, as I mentioned, output is king and good die outer king. So if you're not centered in the process window, maybe your edge die go out a little sooner than they would have in the past. So customers are very focused on working with us to kind of optimize that space, and that's where we're seeing a lot of the pull currently, particularly in the contract business.

Got it. We did touch upon advanced packaging. It is a relatively new dynamic for the industry, right? And an incremental growth driver for Applied. We talked a little bit about this, but you mentioned the EPIC partnerships around next-generation advanced packaging development. You have a number of memory customers. Obviously, I think they're focused on things like HBM.

But you announced TSMC, which is more the 2.5D, 3.5D advanced packaging. Today, the Applied team announced a partnership with Broadcom, who also has a plan to bring in some in-house capability on next-generation advanced 2.5D, 3.5D advanced packaging architectures. So you've got this plethora of different sort of topologies for packaging, in which case you have a big portfolio of tools that supports all of these like -- how does all of this drive service intensity for the package part of the business from an AGS perspective? And is this a new sort of facet of growth for the AGS team as well?

I think it is a new opportunity for us. I mean we're #1 in packaging in HBM, and we've got a great position in 3D SoIC in particular. And those are the 2 fastest-growing portions of the packaging market as we see it. So as those tools go into factories, it creates a new opportunity for us on service.

Historically, there are sort of front-end tools and back-end tools, and there wasn't a lot of overlap between the 2 of them. With heterogeneous integration, advanced packaging, we're seeing the front-end requirements and that complexity, the sensitivity is moving to the back end. So those customers are looking for support from Applied on the service side because we've got the understanding of how those tools need to be maintained in an HVM environment. So I think we're in the very early stages of this as we see going forward. But I do think it creates an incremental opportunity for us in the service business in addition to the great opportunity that we have on the equipment side of the business.

In addition to the strong revenue growth that your franchise delivered in the April quarter, gross margins improved 30 basis points sequentially, 120 basis points year-over-year. Operating margins grew 100 basis points to 29%. That's the highest level in over 2 years, right? And on the gross margin improvements, how much of this is more attach of some of the higher value-added services offerings like AIx, software suite, remote monitoring, which I assume is gross margin accretive. And on the strong operating margins, you're within striking distance of 30%. And I know the team has historically thought they could drive AGS operating margins longer term into that sort of low 30% range. Is that still kind of the way we should think about the long-term operating margin construct?

Yes. We think about it 2 ways. And you kind of commented on them in your question. I mean the top line growth is really all about value creation for customers. The more value we can create, the more good die out, the higher output from the factories, customers are willing to do more service work with us, which we appreciate. And this really requires deep, deep integration with the customers into their factories, really good communication, complete alignment on what needs to be done in the facilities. So as we continue to put things into the service innovation pipeline, those are things that add value. That's primarily driving the top line.

The bottom line growth is really all about efficiency. I mean we have a pretty big operation, $6.5 billion or so operation on a run rate basis. And it's -- that infrastructure is there. And the more that you put through the infrastructure, the more efficient it is. So that's very important. We have continental distribution centers on each of the major geographies. We have 3 of them around the world. They've all been fully automated now with robotics. So where we were touching every part stocking, destocking, packing, now like 60% of that is done automatically.

So it's kind of touch-free until it goes into the box. So that allows us to increase velocity of material through those. It's more efficient from a stocking standpoint for us. And it's better for customers because we can turn things around with higher accuracy and faster.

So we're pretty relentless about looking about how we drive efficiency in the business. And that's really what's going to drive the operating margins for us. So again, we think about it 2 ways. Top line is all about value. Bottom line is all about efficiency, and it's -- in this infrastructure. So as utilization has gone up in our customer facilities, we're essentially -- it's not the exact same infrastructure because clearly, we're adding, but we're driving a lot more material through that infrastructure, so it's more efficient.

I see. And for the final question, I am going to tap into your many years of running the -- being a part of the semiconductor systems business. And now that the team breaks out your semi systems gross margins, it's been a positive surprise for us and for the market, right? Almost 55% gross margins, 35% operating margins for the systems business, significant improvement over the past 2 years, has been driven, I think, by higher value as you bring new capabilities like IMS to the market.

As you look at your pipeline of new solutions, does the incremental system gross margin for your next-generation tools carry higher product gross margins? And what is the team doing to drive better cost efficiencies of your systems, which could also help gross margins for the systems business going forward?

Sure. I think as you pointed out, this is really all about value creation, value creation for customers. When you look at these complex tools like an IMS tool, it's a unique capability for Applied to be able to do this. And so they're very high value for customers.

The other thing that if you just look at the data today, in the last, I think, 3 months, we released 5 really advanced new products, not just minor changes, but quite advanced, Trillium being one of them. There's a selective nitride deposition on the precision tool. There's a treatment tool for Veeva that was part of the Master Class. These are all unique capabilities that are helping customers solve really difficult problems, so they've got high value.

So we think -- as I think Brice and Gary both talked about, long range, as we continue to see more and more of these products moving out as the technology continues to evolve, it's going to be beneficial for Applied overall from a gross margin standpoint.

Perfect. Well, we're just about out of time. Tim, thank you for the update on the high-level overall business of Applied and the great insights on your AGS franchise. Thank you very much.

Yes. Thanks for having, Harlan. Appreciate it.

Welcome to the Applied Materials Second Quarter Fiscal 2026 Earnings Call. [Operator Instructions] I would now like to turn the call over to Mike Sullivan, Corporate Vice President of Investor Relations. Please go ahead.

Good afternoon, everyone, and thank you for joining today's call. With me are Gary Dickerson, our President and CEO; and Brice Hill, our Chief Financial Officer.

Before we begin, I'd like to remind you that today's call includes forward-looking statements which are subject to risks and uncertainties that could cause our actual results to differ. Information concerning these risks and uncertainties is discussed in our most recent Form 10-Q and other filings with the SEC.

Today's call also includes non-GAAP financial measures. Reconciliations to GAAP measures can be found in today's earnings press release and in our quarterly earnings materials, which are available on our website at ir.appliedmaterials.com. In addition, any comments regarding calendar 2026 refer to Q2 of this fiscal year through Q1 of fiscal 2027, which will be a 14-week quarter.

Next, I'll share some calendar items. I'm pleased to announce the second event in our 2026 Master Class series. We'll cover DRAM and advanced packaging on a live webcast on Thursday, June 25, at 9:00 a.m. Pacific Time.

Next, I'd like to remind you about our two special events during SEMICON West. On Monday afternoon, October 12, we invite you to the unveiling of the new EPIC Center in Sunnyvale, California. And on Tuesday morning, October 13, we hope you'll join Gary, Brice and our business unit leaders for our investor breakfast presentation at the Yerba Buena Center in San Francisco. You can join us in person or on a live webcast.

And with that introduction, I'd now like to turn the call over to Gary Dickerson.

Thank you, Mike. In our second fiscal quarter, Applied Materials delivered record revenue and earnings, along with our highest gross margin in more than 25 years. With rising demand and increasing long-term visibility from customers, we see an exceptionally strong foundation for sustained multiyear revenue and profit growth. The momentum in our business is being driven by three key factors: first, the rapid global build-out of AI computing infrastructure. Second, Applied's leadership positions in the most enabling and highest value areas of the market, particularly leading-edge foundry logic, DRAM and advanced packaging. And third, strong execution across our operations and supply chain.

In my prepared remarks, I will provide an update on how our markets are evolving, explain how Applied is enabling the AI road map with our differentiated technology and exciting pipeline of new products and outline how we're transforming the way we work with customers, partners and across the company to drive higher velocity, increase operating leverage and scale more efficiently.

Over the past several months, global AI adoption has continued to accelerate as improvements in the performance and cost of AI computing are translating into real-world applications that deliver compelling returns for users. If I look at our own company as an example, today, we have more than 35,000 AI users across our global workforce. We are deploying AI to drive new scientific breakthroughs, accelerate research and development programs, optimize factory and supply chain operations, increase innovation and productivity and services and automate workflows across our corporate functions. This enables us to redirect resources towards higher value work and grow the business significantly faster than our headcount. Similar dynamics are playing out across a broad range of industries and organizations. Publicly available data indicates that global token generation has increased more than threefold in just the past three months.

Importantly, AI demand is not only growing rapidly, it's also diversifying. Since the beginning of the year, there has been a meaningful increase in agentic applications, which layer on top of continued growth in generative AI training and inference workloads. AI computing architectures are workload-specific and optimized for different generative, agentic or physical AI models. Agentic AI models do more than respond to queries, they plan, reason and execute tasks autonomously. They therefore require a computing architecture that is more CPU-intensive, while also increasing demand for DRAM and NAND. As agentic AI applications grow, they provide an additional tailwind for wafer fab equipment.

Last quarter, we said the availability of clean room space was a key factor pacing the rate of industry investment. As customers find new ways to reallocate or create space, we are seeing incremental requests for equipment deliveries in 2026, and we now expect our semiconductor equipment business will grow more than 30% this calendar year. Given the unprecedented demand environment, we are working closely with our customers on longer-range planning. Our largest customers are providing rolling 8-quarter forecasts, so we can prepare the required manufacturing capacity and service resources for their ramps. With this improved visibility, we see continued growth across this extended planning horizon into 2027 and beyond. And we are investing to support our customers' expansion plans.

For AI computing, leading-edge foundry logic, DRAM and advanced packaging have the greatest impact on overall system performance, power efficiency and costs. As a result, we expect these three areas to account for more than 80% of the year-on-year growth in total wafer fab equipment spending in 2026 and see a similar profile in 2027. These are also the areas where Applied has strong leadership positions and an innovative pipeline of next-generation technologies. In leading-edge foundry logic, Applied is the clear #1 process equipment provider with highly differentiated solutions in materials deposition, modification and treatments, conductor etch and e-beam technologies. Gate-all-around nodes grow our available market considerably while also providing a catalyst for multiple points of market share gains. This quarter, we announced two new products that further strengthen our gate-all-around portfolio.

To meet the requirements of different AI workloads in the data center and at the edge, chip makers provide designers with a range of transistor options with some tuned for peak performance and others turned to use the lowest amount of power. This tuning is achieved through precise optimization of the materials in the metal gate stack. Our new Trillium ALD integrated material solution precisely deposits metals in the most complex gate-all-around transistor gate stacks. By integrating multiple metal deposition steps in a single platform, Trillium ALD provides angstrom-level thickness control of metal gate stack layers giving chip makers maximum flexibility to tune threshold voltages across different transistors.

In advanced foundry logic, shallow trench isolation or STI, is used to electrically separate neighboring transistors. These narrow isolation trenches are some of the smallest structures in a gate-all-around device. Our new precision PECVD system uses an industry-first selective bottom-up deposition process to place material exactly where it's needed and protect the STI structure from damage during subsequent processing steps. By preserving the original shape and height of the isolation trench, our new PECVD solution reduces parasitic capacitance and lowers leakage to boost device performance. In advanced packaging, Applied is also the overall leader with strong positions in high-bandwidth memory and 3D chiplet stacking. We expect to grow our packaging revenues more than 50% in calendar 2026 and are very well positioned at upcoming packaging inflections.

We recently announced our intent to acquire NEXX to further strengthen Applied's portfolio of panel-level technologies, which are designed to enable larger body packages for AI accelerators. In DRAM, AI computing is driving incredibly strong demand and customers are aggressively adding capacity at 6F squared nodes, while accelerating their development of next-generation device architectures. Applied is the #1 process equipment provider in memory today thanks to our very strong position in DRAM wiring, patterning and peripheral logic steps. We expect to gain additional DRAM market share at upcoming transistor and device architecture inflections. We will provide more details at our next master class in June, which will cover both our DRAM and advanced packaging technology road maps.

To accelerate the industry's road map further, we are changing the way we work with our customers and partners by creating a new model for collaboration and innovation. Applied's global EPIC platform is designed to significantly reduce the time it takes to commercialize breakthrough technologies all the way from early-stage research to full-scale manufacturing. For chip makers, EPIC provides earlier access to Applied's R&D portfolio and faster cycles of learning through the co-location of key innovators from customers, partners and Applied. In addition, EPIC co-innovation programs will provide us with greater multi-node visibility to guide R&D investments and resource allocation, increase R&D productivity and value sharing and accelerate design wins for applied equipment and services.

The centerpiece of the platform is the new EPIC Center in Silicon Valley which remains on track to begin operations in the fall. Earlier this week, we announced our EPIC co-development engagement with TSMC, who joined as a founding partner together with Micron, Samsung and SK Hynix. We're also excited to announce our first 3 EPIC university partnerships with ASU, RPI and Stanford as well as the development partner agreement with Advantest.

We are finalizing additional EPIC agreements that we will publicly announce in the coming months. To accelerate the transfer of new technologies from Applied's labs to customers' fabs, we are also driving new innovations in service. Service is another important growth driver for Applied as we increase the revenue we generate per tool on top of a growing installed base. As a result, we expect Applied Global Services to deliver a sustainable annual growth rate in the mid-teens and potentially higher this year.

Our advanced service solutions enable customers to accelerate production ramps and optimize output, yield and cost in their high-volume manufacturing environment. Today, we have more than 35,000 chambers connected to our proprietary AIx software capabilities that use AI-powered monitoring, diagnostics and analytics.

Before I hand it over to Brice, let me briefly summarize. AI adoption is accelerating and diversifying, fueling broad and durable demand for semiconductors and semiconductor equipment. Leading-edge foundry logic, DRAM and advanced packaging are the most critical drivers of performance, power efficiency and cost for AI computing. As a result, we expect these three areas to account for more than 80% of the year-on-year growth in total wafer fab equipment spending in 2026 and see a similar profile in 2027. These are areas where Applied has strong market leadership today and a high-value pipeline of next-generation products, providing us with an exceptionally strong foundation for sustained multiyear revenue and profit growth.

Finally, through EPIC, our advanced services portfolio and rapid AI adoption across our operations, we are transforming how we work with customers, partners and inside Applied Materials, driving higher velocity, improving value capture and efficiently scaling the company as the industry grows. Brice, over to you.

Thanks, Gary. AI is driving wafer fab equipment spending to the areas where Applied has been investing the most and we see the positive mix continuing in the second half of the calendar year and well beyond. As the market shifted to our areas of strength in Q2, we delivered double-digit sequential and year-over-year growth across revenue, operating profit and earnings per share.

I'd like to thank our teams and partners for making these results possible by meeting strong customer demand for our AI enabling materials engineering technologies and systems. On today's call, I'll provide an update on the demand environment, discuss several strategic priorities, summarize our Q2 results and provide our Q3 guidance. Since we spoke February, the demand outlook has strengthened across almost every leading indicator we track. Cloud service providers continue to increase capital investments. Most leading-edge logic and DRAM fabs are running at full capacity. Our customers have announced more fab projects and are giving up the clearest and longest visibility we've ever had. Customers have been using a variety of techniques to increase clean room capacity this year, which is growing the market and our revenue expectations. And based on our latest discussions with them, we expect 2027 will be another strong record year for the industry.

Next, I'll summarize several of our strategic priorities. Our top priority is increasing output to serve our customers' growing demand. We've nearly doubled our manufacturing capacity to support them with expansions in the U.S. and Europe and an additional new manufacturing center in Singapore. We've increased our build plan, inventory positions and logistics capacity. We are systematically translating our 8-quarter customer demand forecast into a consolidated signal to our suppliers. Ensuring they have the visibility they need to make their own capacity and resource additions.

Next, I'll share how this translates to profitability, funding our collaborative R&D process, helps us identify the highest value technology challenges and gives us line of sight to the most compelling solutions. As we bring newly developed tools to market, our portfolio becomes more valuable and our gross margins expand. In fact, our non-GAAP gross margin has increased 800 basis points since Gary became CEO in 2013. It is now crossing 50% at the company level and approaching 55% in Semiconductor Systems. In addition, we are focused on driving higher operating profit and leverage with productivity tools and plans being deployed across the company. While fully investing in the R&D that grows our business and gross margins, we also expect to increase spending more slowly than revenue and deliver increasing operating profit.

Next, I'll summarize our Q2 results. We generated record revenue of $7.91 billion, which is up 13% sequentially and 11% year-over-year. Non-GAAP gross margin reached 50% in the quarter, increasing 80 basis points year-over-year driven by value-based pricing from our most differentiated products, coupled with ongoing manufacturing cost innovations. Non-GAAP operating margin expanded to 32.1%, up 140 basis points year-over-year. And we delivered record non-GAAP earnings per share of $2.86, which was up 20% year-over-year.

Turning to the segments. Semiconductor Systems delivered record revenue of $5.97 billion, which was up 16% sequentially and up 10% year-over-year. The transition to gate all around nodes, along with capacity additions at leading-edge FinFET nodes drove record foundry revenue as well as record revenue across ALD, epitaxy and materials treatments. DRAM revenue of $1.7 billion grew 18% year-over-year. Advanced packaging revenue is accelerating this calendar year within both foundry logic and DRAM, and investments are shifting toward our leadership positions in 3D stacking. Segment gross margin and operating margin both increased year-over-year.

Applied Global Services delivered record revenue of $1.67 billion, which is up 17% year-over-year, reflecting the benefit of higher fab utilizations. Services growth remained strong as our installed base expands and customers choose our most advanced services to boost output and yield. AGS also generated year-over-year increases in gross margin and operating margin. Other revenue of $280 million was in line with our expectations. China represented 24% of our semiconductor systems plus AGS revenue. We expect our business in China and our ICAPS business worldwide to be flat to slightly higher in the calendar year. Cash from operations was $845 million. Capital expenditures were $635 million, resulting in free cash flow of $210 million. We distributed $765 million to shareholders, including $365 million in dividends and $400 million in stock repurchases. In March, we announced a 15% increase to the quarterly cash dividend and achieved a goal we set several years ago to double the dividend per share.

Now I'll share our guidance for Q3. We expect company revenue of $8.95 billion, plus or minus $500 million, which is up nearly 23% year-over-year. We expect non-GAAP EPS of $3.36 plus or minus $0.20, which is up nearly 36% year-over-year. Within this outlook, we expect Semiconductor Systems revenue of around $6.9 billion, AGS revenue of about $1.75 billion and other revenue of around $300 million. We expect non-GAAP gross margin to increase modestly to approximately 50.1%, and we expect non-GAAP operating expenses of around $1.485 billion. Finally, we're modeling a non-GAAP tax rate of around 11%.

In summary, the growth in AI that we've been investing for is now in full force. As a result, the industry spending mix has shifted to leading-edge foundry logic, DRAM and advanced packaging where Applied has built the #1 process equipment market positions. We are investing with confidence to support the strong long-term growth our customers are giving us visibility into and ensuring our suppliers do the same. Finally, we are using the benefits of the AI technologies we enable to accelerate innovation and revenue generation and increase operating leverage and shareholder returns.

Now Mike, please begin the Q&A session.

Thanks, Brice. To help us reach as many people as we can on today's call, please ask just one question and no more than one brief follow-up.Operator, let's please begin.

And our first question for today comes from the line of C.J. Muse from Cantor Fitzgerald.

You talked about 8-quarter rolling visibility with customers. And I'm curious how that might be causing order patterns to change, perhaps upfront payments, a change in the pricing environment. I would love to hear how you're thinking about your relationship with customers, given how tight equipment is and will likely be for years to come.

Hi, C.J., it's Brice. Thanks for the question. So yes, for the large customers, we're definitely working on 8-quarter rolling visibility with them. That helps us primarily plan the supply chain. Our supply chain needs that kind of lead time in order to make their investments and expansions across the large group of suppliers. We've said before, we have on the order of 2,000 direct suppliers and a large number of components for each tool. So that's the primary reason. From a payments perspective, as you know, some of our customers do -- we do require deposits, but that's not across the board. On the pricing side of things, our pricing generally works as long-term contracts with customers based on projects, so you can picture it as a 2- to 3-year pricing contract for a particular project.

So pricing moves relatively slowly in the environment. What has been driving our pricing up over the past couple of years and helping us grow our gross margin over the past couple of years has been the gradual enriching of the portfolio. So every time we launch a new solution, it's typically a more important solution in terms of its value to customers, solving complex problems. That makes the portfolio stronger, and we price accordingly for that. And that's why you've seen our systems gross margins improve over the last 2 years. So I'll leave it at that. No -- at this point in time, no change in the overall model.

Yes. C.J., this is Gary. I would say relative to pricing, really, we're in a great position in the industry in that the most critical innovations for AI computing are in the sweet spot for Applied in leading-edge foundry logic DRAM in advanced packaging and really applied innovations or enabling these architecture inflections that are providing tremendous improvements in computing. So that puts us in a good position with tailwinds to drive pricing in the near term. And then also, as we're creating more value for our customers, it creates an opportunity for us to capture more value from Applied Materials. So I have high confidence that the margin progress that we've seen over the last few years is going to continue as we go forward.

As a quick follow-up, just on the gross margin side, a great guide with, I'm assuming, depressed China. How should we think about gross margins beyond the July quarter, particularly, as it appears as silicon will continue to grow sequentially well into 2027 and beyond?

Hi, C.J., thank you. So we guided 50.1% for the company-level gross margin. I'll just remind investors that we're now reporting our gross margin for each operating segment. So they'll be able to see that our Semi Systems gross margin is 54.8% in Q2, and that represents the improvements we're talking about. And on the systems, to your point, we expect to continue to make improvements, that's really driven by the gradual strength in portfolio as we launch new equipment. So the improvement will be slow, but we expect continued improvement in the gross margins. And that's really where you get that guide of 50.1%. So a slight increase for Q3 from Q2, and we would expect to be able to continue that process going forward as we launch new tools and solutions.

And our next question comes from the line of Stacy Rasgon from Bernstein Research.

For a 30% year-over-year equipment growth, so that looks like you're suggesting something like $14.5 billion, $15 billion in the second half of the calendar year for equipments? And I guess, I just want to verify that's true. And I guess what are your thoughts -- it was too early about calendar '27, but I mean given '26 is still a constrained year. I guess what are your thoughts on WFE growth versus your own growth in '26? And is there any reason to think that '27 couldn't be even better as cleaners become more available? I guess just how do we think about the trajectory of growth into '27 as the clean rooms like come online relative to what you're suggesting as we see into the back half of this year?

Yes. Stacy, good to hear from you. So on the 30% year-over-year equipment growth, it does suggest a second half on the order that you described. We do think that the demand signal is very strong. Our customers increased orders in the last 90 days, as Gary described, as customers are finding new floor space and clean room solutions. And I think investors should think about when they think about the road map of capacity. We've said before, we're tracking over 100 factory projects globally. We added more than 10 just in the last quarter. So you should expect a pipeline of new clean room coming on board, so customers can continue to grow the wafer starts across the end markets that we're talking about. So we do expect growth in WFE. We do expect growth in wafer starts overall, and we think that's headlined by the AI demand function.

Yes. Again, as Brice indicated, '27 still looks like a strong growth year for us. And I can tell you that I'm having constant conversations with customers, and they're looking at '27, and they're looking at '28. Again, because the compute demand is growing so quickly. Earlier on the call, we talked about agentic AI on top of what we've been seeing and that we're modeling incremental CPU demand, incremental DRAM and NAND demand, that's a meaningful increase on top of what we had been forecasting previously. And then you go beyond that with physical AI coming in the future, we think this demand environment is going to continue strong for a number of years. And certainly, as I'm having many conversations with customers, that is a major focus for them.

Got it. That's helpful. For my follow-up, I wanted to ask about your manufacturing capacity. So you see you doubled it. So is it all ready to go? Like how much revenue -- I guess how full -- maybe the question I'm asking is like if it was full, how much revenue capacity could you actually support, and is there actually more margin upside as that capacity that you brought online actually starts to fill up?

Stacy, yes. So on the first part, from a floor space perspective, it's ready to go and available. We would have to fit that up and hire people as we ramp. So we have the capability of significantly expanding the output. Last quarter, I said we could double as we're growing, it eats into that a little bit, but that's a rough approximation. So we have significant capacity available. Our job is to work with the supply chain and have the supply chain work at the same speed we are.

And our next question comes from the line of Vivek Arya from Bank of America Securities.

If we look at WFE growth this year, it's very strong, but it is still the semiconductor industry growth, mainly because semis are being driven more by pricing rather than units. So Gary, I'm curious that as you look over the next 1 to 3 years, do you think that semi industry growth comes more from units, or it still continues to be reliant on pricing?

Yes. Thanks for the question, Vivek. Again, when I'm talking to our customers or our customers' customers, I think everyone sees tremendous increase in computing demand going forward. And I talked about all these layers of demand that are increasing as we go forward. So I think that, and as I mentioned also with customers, the focus -- I was just with a customer 2 days ago, they were worried about the supply all the way into 2030. So again, the computing demand, I think, is increasing. I'm not going to speculate on pricing.

What we do see, again, is broadening in terms of the customers that we're working with because everyone sees this increase in compute demand, and that is really great for Applied because all of the markets that are growing the fastest, the leading-edge foundry logic. We're seeing a broadening of demand, very strong demand for DRAM and high-bandwidth memory. The ecosystem working on new packaging architectures that drive major improvements in performance and power tokens per second per watt. So again, all of this is great for Applied. I'm not going to speculate on the units versus pricing. But what I would say is that the market environment for Applied has never been better and our positions in the market have never been better.

And for my follow-up, if your system sales are growing over 30%, and there's a lot of discussion that maybe growth accelerates next year. How should we think about growth in AGS given that level of growth? Is there some kind of correlation? Is there some kind of kind of lagging growth that we can start to see acceleration even in AGS at some point?

Hi, Vivek, it's Brice. Yes, thanks for that question. So we had previously communicated low double-digits growth expectation for AGS, our services business. And to your point, as the systems business is running higher, and we've raised expectations for the systems business, that means the AGS installed base will grow. So the service opportunity for AGS grows and as well as the spare parts and as factory utilizations have improved. So we are going to raise our expectations for AGS to mid-teens for sort of the normal environment. And we think this year will end up being a little bit higher than that because we have so many -- so much utilization improvement across the industry and so many new factories that, that gives extra bump. So I would say mid-teens from a medium-term modeling perspective and higher than that this year.

The other thing I would say on AGS, there's never been a time where output yield innovation is more valuable than today. And I mentioned earlier on the call that we have over 35,000 chambers connected to our AIx servers. So we're also driving service innovations with AI-enabled applications, predictive models to improve wafer fab output, yield and cost. So I think in addition to the growing equipment business that will help accelerate AGS growth. We're also driving higher value service innovations that will also support that growth.

And our next question comes from the line of Timothy Arcuri from UBS.

Brice, I wanted to go back to this up 30% -- at least 30% this year as well. So if I just do the math, I mean, it sort of implies that you barely grow off of the July levels, which doesn't seem possible given you're booked out for at least a year. So I would think you can probably grow at least the same number of dollars in Systems that you've been growing in the last 2 quarters. So that would put you like more above 40%. So I'm surprised you said above 30% and not above 40%. So are you just being conservative, or is there something like in the back half of the year where the sequentials can't grow as fast from an absolute Systems level. Can you just talk about that?

Sure. Just to reiterate, we're saying 30% growth year-over-year in our Systems business. And Tim, I think we're not giving explicit linearity for the out quarters, but we'd be comfortable if you just assumed it was linear from Q3 to Q4 to our fiscal Q1.

Okay. So it's going to grow about the same on a dollar basis? Brice is that -- sorry, just to clarify.

On a dollar basis, I would just draw a line and say it grows linearly from our Q3 guide through Q1.

Okay. Got it. And then, Gary, can I ask you about Huawei story, I mean, how do you think about -- I mean that letter in and of itself is not -- like it doesn't really hurt you at all. But how do you think about the risk of that broadening because there's a lot of examples of these fabs being co-mingled in these clusters, some of which you can ship to and some of which you can't. So is there a risk that this spreads and maybe it sort of then captures more in the net, where you can't ship to these fab complexes that have stuff that's allowed and stuff that's not allowed co-mingled in the same complex. Can you just talk about that?

Yes. Thanks, Tim. Let me just clarify. Relative to our semi equipment business, we'll grow 30% or more in the year. And it's really supply chain. That's one thing that is, I think, an issue for everybody. As Brice mentioned earlier, our operations can scale significantly beyond where we're at right now. But it takes time for the supply chain to respond. And certainly, we're working that as rapidly as we can. And I really do believe we've made tremendous improvements in our supply chain and our operations over the last several years.

So I feel really good about that. Relative to restrictions, I don't really want to comment on that. All of that has been factored into our guide for the quarter in our viewpoint on growth in the calendar year. And again, what I would say, if you look at the mix of business going forward, the AI compute innovations are really going to drive in '26, '27 and going forward, leading edge foundry logic, DRAM, high-bandwidth memory and advanced packaging. And those are the areas where Applied has tremendous strength, and we are absolutely positioned to gain share as these inflections happen in all of those different markets.

And our next question comes from the line of Harlan Sur from JPMorgan.

In addition to technology migration and greenfield capacity build-outs, your customers are looking at their existing capacity footprints. They're trying to figure out how they can squeeze more wafers out of their fabs, it's more good die per wafer. So focusing essentially on throughput and/or focusing on yield or both, right. Is the Applied team seeing this? And is this an incremental driver of growth, more tool sales to alleviate bottlenecks maybe upgrades to improve tool throughput or maybe advanced services and analytics adoption to improve productivity and yields. Is this driving some of the incremental growth?

Yes. Thanks for the question, Harlan. Yes, absolutely. And I mentioned this a little bit earlier. Output and yield innovation is really a key focus for all of our customers. Again, it takes time for them to create more floor space to ramp their factories. So there is a huge focus and engagement, and this is really helping us to drive our service growth rate at a higher pace than we've seen in the past or even that we were anticipating in the past. And that's where I would also come back to how we're implementing AI inside Applied Materials. We have all of these connected chambers. Most of them are connected remotely, so we can have instantaneous experts connected into all of those different chambers. And again, this is elevating, increasing our expectations for our service growth rate going forward.

And then maybe for Brice, in addition to the strong growth in AGS, gross margins improved 30 basis points sequentially 120 basis points year-over-year. Operating margins grew 100 basis points sequentially to 29%, right? I think that's like the highest level in like the past 2 or 3 years. On the gross margin improvements, how much of this is just more attach of some of the higher value-added services like your AIx software suite, remote tool monitoring solutions, which I assume are all gross margin accretive. And then on the strong operating margins, I did notice that your OpEx for that segment was down about 8% sequentially. Was that just cost efficiency activities, or does that delta come back this quarter?

Thanks for the question. So on the gross margins there, I think the mix to more transactional and more spare parts in the quarter have helped us. So that's one of the drivers. You're right, the new service products that we're providing also help from a margin perspective. So I think those are the major drivers. And I think on the expense side, we did some restructuring that helped us with expenses. So -- but I would expect that we're continuing to invest in that business. We're building a training center. We're adding to our customer engineers and essentially expanding for the ramp at this point. So no change in expectations from a spending perspective. I think the margins are at a healthy spot at this point.

And our next question comes from the line of Atif Milk from Citi.

Gary, with your acquisition of NEXX in panel level packaging, can you help us understand how does that fit with your JV with Besi, what is the adoption time line for large panel packaging?

Thanks for the question, Atif. So I would say that we are the leader in packaging. I mentioned earlier on the call, we're going to grow that business over 50% this year. And I would say that in the entire semiconductor industry, this is one of the most exciting areas driving compute architecture innovation. So how you connect computing components together in large body sizes is an enormous focus for everybody. So at Applied, we have been investing. You mentioned the hybrid bonding technology and the acquisition of NEXX, that's a leading supplier of large area packaging, electroplating. There are a number of different innovations that we're driving. And this fits into that overall strategy where we also have a full-flow packaging EPIC Center, that is the only one in the industry.

We're engaged with all of our existing customers and then also there are a number of people working on new large body sized architectures again, to really drive tremendous improvements in performance and power the tokens per second per watt. So this is all part of that overall strategy that we've been investing in the last few years. You see, again, over 50% growth this year in our packaging business. Next year, we anticipate also a strong growth.

And the timing for those architectures, I would say that there is a tremendous amount of focus to bring those architectures to market as fast as possible. I'd really rather not comment on specific timing. I would just say that Applied, we have deep connectivity across the entire ecosystem with all of those people that are driving those new architectures. We have the broadest packaging portfolio in the industry. And we are really at the foundation of the key innovations needed to enable those new architectures.

Great. As my follow-up, Brice, I understand you guys do not comment on a particular project. But you said there are 10 new projects in the pipeline, and there's a lot of investor interest in Terafab. Will that be a project that will be in the pipeline, or is that more like next year?

Well, just to correct the record, I said more than 10. And you're right, I can't name any -- I can't be specific about any of those, sorry.

And our next question comes from the line of Krish Sankar from TD Cowen.

Gary, the first one for you. Applied had an interesting slide at SPIE that showed DRAM 1 delta node to be more litho intensive compared to 1 gamma. And obviously, you have 4F squared coming after that, which is more depth and etch intensive. So I'm wondering, do you expect 1 delta to be a big node and if so, could that slow down the depth and etch intensity in DRAM on an interim basis?

Yes. Thanks, Krish, for the question. So let me -- from an overall standpoint, DRAM, we are the #1 process equipment provider. And there's innovation happening at 6F squared, 4F squared and certainly going forward to 3D DRAM. So if I look at Applied in terms of near-term innovation, periphery CMOS logic being upgraded for higher performance and lower power transistors is a strong driver for our DRAM business, including very strong growth in epi. Of course, in DRAM, HBM packaging is a huge focus for everyone, stacking more chips vertically. And then for Applied Materials deposition with wiring and patterning, and we have very strong positions, leadership in conductor etch and e-beam technologies in DRAM.

So what I would say is that there is a lot of innovation happening in 6F squared that really extends our process equipment leadership and is contributing to the growth that we're talking about on the call here today, and we're even better positioned for 4F squared and 3D DRAM. I think if you remember, over the last a little more than 10 years, we gained 10 points of overall share in WFE and DRAM. We've been really performing incredibly well over the last few years and then this year is another great year for our DRAM business.

So I think for Applied, we're in a great position. And relative to litho intensity, I believe that going forward, for sure, materials intensity will increase, especially with 3D DRAM, and we're in a great position with 6F squared innovation, 4F squared and 3D DRAM.

Gary, just a quick follow-up either for you or Brice. I understand your customers are giving visibility into 2028. I understand for your capacity and personal planning purposes, but also some new greenfield and NAND scheduled for 2028. Is it fair to assume '28 could be another growth year for WFE after impressive '26 and '27, is it too early to make that call?

We don't think it's too early, Krish. I think we think it's secular growth. AI is the headline workload that's driving the secular growth. So we think you'll see more and more capacity plans from customers. That's the expectation is right now, it is factory limited as we've said, but we expect customers to continue adding projects to grow the outlook.

And our next question comes from the line of Shane Brett from Morgan Stanley.

My first question is, so the Gartner data showed that your conductor etch market share picked up about 300 basis points in 2025 despite weaknesses in other areas of your portfolio. Apologies if this is a bit of a trailing, backward-looking question. But how should we think about where those share gains came from? How should we think about your path towards being a market leader in this space where are those sort of incremental share gains going forward going to come from for your conductor etch business?

Yes, Shane, thanks for the question. So etch, let me talk about going forward, and then I'll talk about the overall market environment. So this year, etch will be one of our fastest-growing businesses this calendar year. and Applied as #1 in conductor etch for leading edge foundry logic gate-all-around nodes and #1 in conductor etch for DRAM. And both of those areas I talked about earlier on the call are some of the fastest-growing certainly in '26 and going forward. So we've been building a stronger position in leading-edge foundry logic. We've always been very strong in DRAM. .

And in our products, Sym3 is the fastest ramping product in Applied's history. We recently talked in one of our master classes about our new Sym3 Z platform. We've seen strong adoption there with more than 250 chambers, multi-hundreds of millions of dollars of growth in etch. So I think we're really, really in a great position. As I said, one of the fastest-growing businesses this calendar year. The other thing is really the opportunity for us to co-optimize across our whole portfolio.

So when we're working with our customers on new architectures, whether it's gate all around or wiring or new DRAM architectures, really that opportunity to co-optimize helps us -- we have great technology, but also we have great integration capabilities that's helping us grow this business at a high rate.

Great. And for my follow-up, this may be a little bit critical, but your process control market share continues to decline. Contrary to conductor etch where you are seeing share gains. My question is just how do you see the strategic importance of your process control business within your broader franchise?

Well, I believe this is one of our best opportunities overall within Applied Materials. So last year, we grew this business at a high rate. This year, PDC is going to be one of our fastest-growing businesses. And I'm very, very positive on the growth in '26, and I'm even more positive about our growth going forward. We have clear leadership in eBeam with our Cold Field Emission technology that gives us the highest resolution and the fastest imaging. This year, our optical inspection business is also on track for strong growth. We have a great pipeline of new technologies that we're going to introduce over the next couple of years that will also drive significant growth in PDC.

So one of our fastest-growing businesses this year positioned for strong growth in '27 and as we go forward. And the other thing I would add is there's great synergy with our eBeam leadership with the rest of Applied Materials in accelerating learning rate. So our process equipment teams, they're using this leadership in eBeam imaging to optimize processes at a faster pace. So that is helping us to drive growth in our process equipment business on top of PDC being a great growth business in and of itself. So again, I don't share your view on the perspective on the business. I think this is a great business, one of our fastest growing this year, and I'm very optimistic we're going to keep this growth going, going forward.

And our next question comes from the line of Joe Quatrochi from Wells Fargo.

I was wondering if you could talk about your supplies position for greenfield versus upgrades. I would assume given that the portfolio depth and just the number of new greenfield fab projects expanding into next year. How do we think about just the accelerating growth opportunity for Applied relative to peers?

Joe, it's Brice. Yes, I think most of the projects we track are greenfield projects, especially on the logic side. We do very well on greenfield. That's probably the best intensity from Applied perspective. We also do well in upgrades. The one place that that's different is NAND. We don't participate as much in upgrades. But I think what you're seeing in the environment is just a continued search for more and more floor space and more equipment being put in place. You see that in DRAM, you see that in leading logic. And of course, that's been the case in ICAPS as new -- significant new capacity has been added over the last several years. So I think a lot of that investment is greenfield.

And then as a follow-up, increasing the AGS long-term growth expectation. Can you help us just kind of bridge the gap? I think the prior target had included the 200-millimeter business that's now moved into Systems. So I guess like how do we think about the impact of that to the increase in the growth expectation?

Yes. Joe, so that should be a clean look at this point. So we've moved the 200-millimeter equipment business and restated our financials or recast our financials. So that's all in our Systems business at this point. So when we look forward on AGS, we're only talking about the services business. And you may have heard, I just updated our outlook for that and said, mid-teens for the AGS business is sort of a good multiyear growth rate assumption. And then we expect to be higher this year because we have a increase in utilization across the ecosystem and an increase in the number of new fabs ramping relative. So there's a step up this year that helps that business be higher than the mid-teens.

And operator, we have time for two more questions today, please.

And our next question comes from the line of Jim Schneider from Goldman Sachs.

Your mix of business this quarter pivoted quite strongly again to foundry logic. I'm curious, as you look into the back half of this year or whether you see more of a return to spending in greater concentration to DRAM? And then on a preliminary basis as you looking into 2027, which do you expect to be the fastest growing technology area?

Thanks, Jim. Yes, in the second half, it goes back to the theme that we've been describing. You're going to see strong growth in leading-edge logic, in DRAM, in advanced packaging and actually NAND also. So you're seeing a pull really from the headliner AI across the entire end market grouping with the exception of ICAPS. So ICAPS goes back to -- there'll be some digestion with all the capacity that's been put in place the past couple of years. So -- and we didn't call a fast grower between DRAM and leading logic. I think both of those will be accelerating in the second half.

And then as you look at your customers' longer-term plans, can you share where they're seeing the greatest upside to the longer-term forecast across those same areas?

That's a good question. I think we haven't answered which one between DRAM and leading logic will be the most growth across the multiyear horizon. We think both of those will be very strong and actually advanced packaging goes right with it. On the NAND, we see -- we did raise our bit growth forecast for NAND. So probably a few percentage points and at least our view is, it will still be satisfied by upgrades. So we don't expect a lot of new wafer starts on the NAND side, but we do think the demand forecast for bits has gone up on NAND.

Yes. Jim, just also part of the question you asked earlier, we see a similar profile in '27 to what we're seeing in '26. And really, that's being driven by AI computing. The markets that are growing the fastest in '26. We think they continue to grow at that same rate as a percentage of the mix in '27 and going forward.

And our final question for today then comes from the line of Melissa Weathers from Deutsche Bank.

I just wanted to touch really quickly on the ICAPS side that, Brice, you kind of just talked about, but I think if we look at a lot of the analog reports, this earnings season, it does seem like inventories are starting to lean out and maybe utilization of those fabs are starting to get better. So I realize it's not the fastest-growing part of your business. But any updated outlook you guys have on maybe when we could see some more material spending on that core ICAPS business going forward?

Melissa, good question. And that's a great point. There are some strong areas in ICAPS. I think analog and photonics, power chips are all areas that we're seeing that are very strong. So that makes sense to us. I would just encourage investors to think ICAPS is still a very strong market. It's one of our largest markets. I think this is the first year in a few years where leading logic is actually ahead of ICAPS, but ICAPS is still plugging along, still adding capacity at a significant rate. It will just be -- it just won't grow a lot year-over-year until we digest the capacity.

The utilizations have improved in ICAPS. So to your point, I think 300-millimeter utilizations are in a very healthy range at this point. So when we look forward, we expect ICAPS from an equipment perspective, to eventually start growing at the same rate the devices are, which is mid- to high single digits. We just have to get past this period where utilizations catch up with the installed capacity.

Got it. If I can squeeze one last one in, in the end of the call. Just the EPIC Center, I know you guys are a couple of months away from unveiling that. So just any last couple statements you can make on what that could do to drive innovation. And I don't know if it would be share gains or competitive advantage, but anything on EPIC Center we should look forward to?

Yes. Melissa. I'm really excited about the EPIC Center. I mean, again, all of our customers and the entire ecosystem earlier is in a race for leadership or AI computing. That is really driving our businesses in leading-edge foundry logic, DRAM, high-bandwidth memory and advanced packaging. And really, when you think about architecture innovation, Applied's material innovations are really at the foundation of all of those architecture innovations across all of those fast-growing segments. So having our customers and partners co-located with these technologies that are enabling the architecture innovations here in our EPIC Center and co-located with our innovators is going to be an acceleration for Applied and for the entire industry.

We've announced 8 partners that we talked about earlier, TSMC, Samsung, Hynix, Micron, Advantest, ASU, RPI, Stanford, and we will be adding more soon. For Applied, it gives us better multi-node visibility because, again, we're working many generations out in the future with all of these different companies and really co-innovating. We are working with them to create these new architectures in the EPIC Center. And it also enables us to be designed in with our equipment and our advanced services for those new architectures. So I'm really excited about the EPIC Center. I think this is the right strategy at the right time, and it's going to be an accelerator for our customers and for Applied Materials.

Well, thank you, Melissa, for your questions. And now, Brice, would you like to give us your closing thoughts?

Excellent. Thanks, Mike. We're really pleased with the investments we've made, puts us in a great position as AI drives incremental demand across the industry. We're in a strong position to grow revenue, expand margins and increased operating leverage. We have a number of upcoming events to call to your attention next week. Tim Dean from our services business will be at the JPMorgan conference in Boston. The following week, Gary will be at the Bernstein conference in New York. And a week after that, I'll be at the BofA conference in San Francisco. We hope to see many of you at those events. Mike, please go ahead and close the call.

All right. Great. Thank you, Brice. And we'd like to thank everybody for joining us today. A replay of the call is going to be available on the IR page of our website by 5:00 Pacific Time, and we would now like to thank you for your continued interest in Applied Materials.

Thank you, ladies and gentlemen, for your participation in today's conference. This does conclude the program. You may now disconnect. Good day.

I think we're set to go. Well, good morning. Welcome to day 1 of Cantor Fitzgerald's Tech and Industrial Conference. This is our second year. Very pleased to have Applied Materials and Brice Hill, CFO of the company. Welcome.

Thank you. Thanks for hosting us. New York, wonderful weather.

Yes, yes. We brought this lovely weather for you. Be assured, it will snow on Thursday.

But to start, I think the world has dramatically changed in the last 3 months, both from your outlook, your competitors' outlook. You called for revenue growth for silicon of greater than 20% growth here in 2026. This is, of course, limited by clean room space availability, resulting in seeing plenty of room for another double-digit year in '27. Do you agree with this line of thinking? And what do you see as the biggest risk to this view?

Well, certainly, the demand signal is extremely strong. I've been talking to people about artificial intelligence and AI. And we've expected our business to pick up based on the demand. When we look at cloud service provider investments -- top cloud service provider investments and data center AI, CapEx are in the $600 billion this year, projected to be in the $700 billion next year. And we've seen a lot of pull from our customers for more advanced logic, more DRAM and more advanced packaging solutions.

Right now, the market is constrained with the leading-edge semiconductor components, those 3 areas. And we expect that demand signal to continue very strong and that dynamic of being constrained not to change in the foreseeable future. That's the expectation.

C.J., you mentioned that it's constrained by fab space. We did, of course, increase our expectations for this year. So there was more space available to utilize this year. But to your point, it's -- we think it's fully utilized at this point with the space available. So the ramp will be metered by space in leading logic and DRAM as we go through the, let's say, next -- it might be more than a year that we experience that.

So with that as a backdrop, are you starting to see customers slot tool availability in '27 and even '28?

We definitely see slotting into '27 at this point, but I'll back up and say foreseeing this ramp that we're talking about, we've been working with customers on improving our visibility to 2 years with all of our top customers. And in fact, all the customers across our network, we're asking for 2 years of visibility into the equipment that they need. And then, we provide that signal to our suppliers.

I've highlighted before that we have over 2,000 suppliers that we work with on a day-to-day basis. You have to give those suppliers a signal if they're going to change their capacity, if they're going to hire people and train people. You have to give them enough lead time to do that because this is a significant increase in output. And so we've got 2 years visibility from our customers. We're providing that to our suppliers. And when it gets closer, when it gets within 1 year, that's part level detail, so the suppliers can truly ensure they can meet your needs.

And in terms of the 2-year slotting, are all customers willing to share that information or only the majority, let's say?

Well, I looked at it this morning, just so I would -- I know that. But I would say it's -- thinking generally about the world, first of all, it's all customers. It's an ask of all customers. I will say that the visibility into the longer list of China customers is more difficult for that period of time. In some cases, they haven't completed those plans. But I think if you think about the largest customers, leading logic, DRAM, NAND, largest customers, those are very complete views.

So if you think about operational readiness, and I know your goal based on that backdrop is to be ready for everything. But this is the first cycle that I've ever seen, and that is a 26-year view, where you're seeing leading-edge logic, DRAM and even NAND all in an upturn. And when you put it all together, you're potentially talking about during COVID, where WFE grew by $35 billion nominal dollars, where this time, it could be $60 billion, $70 billion, $80 billion, $90 billion over a 1-, 2- or 3-year period. And so can the industry support that kind of rapid growth?

We think the industry can support that growth, and that's why the supply chain is so important in this process. As you highlighted, during COVID, we experienced -- there were different pockets of the supply chain that were very difficult to improve. We've never stopped working on supply chain since COVID. We've had to reengineer it to be more regionally self-sufficient. We've had to reengineer it to make sure we have fewer sole reliances across the supply chain.

We've made our own supply chain more global than it was in the past. We've increased our capacity. We've increased our inventory positions, and we've improved the signaling to the suppliers. So I would say at this point, we don't see any roadblocks to the supply chain being able to support the lift that we're talking about.

And maybe a wildcard question that's arisen in the last, I guess, 9 days of war and how important oil production is to helium. I guess, are you hearing anything from the supply chain around challenges there?

I don't have any perspective into our customers. I know from an Applied perspective, we have reviewed that, and we don't think it will be a constraint for Applied at this point in time, but that will be a good question for the customers and how much they use of that particular gas.

Perfect. So maybe moving on to end markets. You obviously talked about a number of growth drivers here in 2026. So maybe to narrow in, for leading-edge foundry, we're seeing a broadening of spend. I guess, maybe if you could speak to that? And is that breadth something that can really help '26? Or is that more of a '27, '28 story?

Well, what you're seeing, I think the strength, the fastest driver are these AI systems. Data center now is about 30% of demand for wafers on the leading edge. So that has just eclipsed demand for PC components on leading edge. We expect it to overtake smartphone components by 2029. So that data center growth on the leading edge is really where the strength of the demand cycle is coming from.

I just looked at a GPU system for 2027 that's being planned. It's got double the GPUs of a previous system. It's got about 50% higher transistor count per GPU. It's got 33% more memory per GPU. So when you tear apart these AI data center systems, you can really see the concentration continuing to improve in all the various compute and memory components.

And when those systems are built in more and more high-complexity solutions, that's what pulls on the advanced packaging techniques. But that leading edge, that is the strongest pull for leading edge. You asked for breadth. I think those 3 markets are all growing, and the market is constrained at this point on leading edge. And so we see 100% utilization across all the leading-edge nodes.

And maybe on DRAM, I guess, can you speak to kind of the dynamics there? Obviously, we're in a supply-constrained environment, but would love to hear kind of how you're thinking about your leverage to HBM. And as we transition to HBM4 and beyond, how Applied Materials' share of wallet should grow?

Yes. I think DRAM similar. It's being pulled by that same demand function. About 15% of the DRAM wafer starts right now are allocated towards high-bandwidth memory. High-bandwidth memory is a double opportunity for the equipment industry because not only are you selling equipment to make the DRAM, the high-bandwidth memory is more intensive from a wafer perspective. It takes 3x as much wafer area to make the same amount of memory with the high bandwidth component. And then, you stack the high-bandwidth memory. So there's additional process steps in order to stack that memory.

We think about 19 extra steps, 15 of those are equipment steps, where Applied Materials provides solutions, and we sell more than 50% of the value of the equipment that goes into those additional packaging steps for high-bandwidth memory. So very strong demand signal. I think the roadmap will continue to build out. Customers are working to accelerate floor space so that they can add capacity. And as new technologies come on, we think there'll be also other stacking and bonding solutions that the industry will develop, and we'll participate in. Those will be inflections that will be valuable for Applied.

And maybe I'm getting ahead on the service question. But I'm hearing upgrades of tools, whether it's even a micron adding an extra layer of EUV to clear up clean room space, replacing KrF tools for higher throughput. Are you seeing that particularly from the DRAM market?

We have. We have heard that -- I think we've heard that both for logic and DRAM, where customers will look to optimize floor space in various ways, and so, not surprising. We're trying to maximize output at this point in time.

And so maybe to combine leading-edge foundry and DRAM and the advanced packaging side of the house, your business has grown from $500 million to $1.7 billion, 2020 to 2024. But last year, it declined while others saw growth. Can you talk to the divergence? And how we should think about the business from here?

Yes. 2024, we're talking about advanced packaging. 2024 was a very strong year. There was the initial build-out of the HBM equipment on the DRAM side. And that was a high number. I think it was about $700 million in 2024. It was a little less in 2025 after that initial build-out. So that slowed down.

As we're looking forward, we put it in the fast lane of growth, C.J., when we talked about semi equipment growing more than 20% this year. The way to think about it is advanced logic or leading logic and DRAM and advanced packaging are all in the fast lane. So they'll grow faster than that 20-plus percent growth rate.

ICAPS and NAND will be on the slower growing side. So you can see the faster groups will grow much faster than the 20% and the slower is much lower. We've said ICAPS, for example, will be approximately flat this year is the expectation.

So back to the advanced packaging, we think it will be fast growing this year. We think there'll be continued investment in HBM, and that will continue to grow. And we think advanced packaging overall will continue to grow with that same AI packaging-driven need.

So Life in the Fast Lane, Eagles then.

Life in the Fast Lane.

So how about on NAND? That's a market that you've talked about is not growing as fast. However, we're obviously in a very tight environment. You've got 1 player adding wafers, everyone else adding layer counts. I guess, why do you think it's not comparable to the other parts of the market?

Well, I think NAND is interesting. The bit demand, the bit rate demand has been very strong and continues to be strong. I think we've said it's in the past that it's in the high teens for a bit for bit growth. The issue with NAND is that the technologies have been so excellent in terms of providing more bit density, each advance in the technology that you just don't need more wafer starts to grow the bit supply by 20% each year. And that's been what's happening.

So if you look at the NAND wafer starts over the last 4 or 5 years, they're relatively flat. We count about 1.4 million wafer starts a month on the NAND side. And that's been roughly flat. So what you see is companies adding capacity for those additional layers, but they don't need to expand the number of wafers out.

We think that continues for the near future. At some point, they'll have to add capacity, but it's not in the near future. So that's what keeps it at a lower growth level from an equipment perspective. But again, there's plenty of NAND demand, the bit demand in the marketplace.

Makes sense. And maybe the last part, ICAPS in China, obviously, a place of concern from investors. And so how do you think about the trends there given your overexposure to both?

Well, it's -- I don't know if I would agree with characterizing it as overexposed to ICAPS. ICAPS is our mature node technologies for Applied Materials. The ICAPS stands for IoT, communications, auto, power and sensor chips. So it's those chips that are built on like 28, 40, 45, 60-nanometer type nodes. We have a very strong position in those nodes.

And if you look over the last few years, we typically have for our Equipment and Services business in the high 20% of the business that goes towards China, and most of that demand has been ICAPS. We had a very small period of time, C.J., where our -- the reported percent of China business for Applied was much higher. And that was when we were serving DRAM. We had a build-out of DRAM capacity in China. So it went into the 40%. But otherwise, I would think, as an investor, 25% to 30% is the reasonable range where we're serving business in China, and all of that is our ICAPS business. So a very important business to us.

We don't think we're overexposed. And we think that business continues. It will be flat over the near future because there's been a significant amount of investment in capacity in China, but it's a very good business for us. A lot of that future growth will be on 28-nanometer. We have an excellent position in the 28-nanometer node. And where we can -- the customers that we can serve, we'll serve -- we'll do a good job of serving. We're also continuing to innovate in that space.

Makes sense. So it sounds like flat is -- means we're derisked. Curious -- I think President Trump and President Xi will have a meeting at some point in the next month or so. And I think there's hope for some sort of grand bargain. And just curious, given the very tight nature of both DRAM and NAND, do you think there could be relaxation of restrictions for perhaps the consumer memory market? Have you heard any inklings of that from Washington, D.C.?

We've heard nothing from Washington, D.C., as far as I know. I have read in the press what you mentioned that some device companies may be trying those memories because of the constraints in the market. Certainly, that would be upside to our outlook if we were able to serve those customers again. The China market has been constrained. That's one of the reasons we grew more slowly. It's been constrained by the trade rules. Currently, U.S. companies are more constrained by trade rules than non-U.S. companies. That's something that we continue to work with the government on. And that would be an upside to our forecast if we could serve those customers.

Perfect. Now, maybe moving to service. You've guided for double-digit growth here in 2026. At the same time, we discussed earlier, upgrades and customers at near peak utilizations. I guess, how are you thinking about the drivers this year? And is there a potential as we layer in maybe more aggressive capacity adds? I don't want to say an uptick to that number, but maybe that number proving to be conservative.

Well, our reported quarter, just reported quarter, our Service business grew 15% year-over-year. Low double digits is our outlook, and it's -- when you think about that business, every single quarter, the installed base of equipment that we can serve from a services perspective grows. We expect the installed base to grow 5% to 10% depending on the market each year. And we're adding service products every single year to our portfolio, which allows us essentially to provide more value to customers.

And finally, the attach rate to our equipment in terms of applying services goes up each year. Customers are building in new cities, new areas. They tend to use the services more because they're looking for qualified labor to help them ramp the factories. So those 3 dynamics help grow that Services business in low double digits.

And certainly, if we have higher utilization or we have more growth on the equipment side, then that grows that installed base and gives us upside to that. But we think that low double digits is a high confidence forecast.

Perfect. And then maybe a part of the business that's not asked much about these days, and you actually just put it into the other category display. I think there's been hope that there would be an OLED-led inflection at some point. Where are we in that? And at what point and what magnitude could that start to help the story?

Yes. OLED, very important. We've got some innovations. We launched a technology called MAX OLED, where we have more of the share of wallet in terms of producing the OLED solutions going forward. And we do think that there will be an inflection with more penetration in the laptop form factor and even larger devices. Television would be the place where you get the most screen size.

Nothing to announce at this point, but we think penetration for OLED will continue to increase across the industry. And we're not reporting that as a reportable segment any longer, but we did say that we would share in investors when the business changes in size. And so we can look forward to that going forward.

And I forgot to mention something on the services side that I want to come back to. We recently did a reorganization. We used to have our 200-millimeter equipment business inside of our Services business. We reorganized and moved that 200-millimeter equipment business and put it in our Semiconductor Equipment business. So when investors now look at our Services business, it is pure recurring revenue. So it is the spares you need for existing tools, and then, the services that we sell along with those tools.

And that gives us -- when we talk about that low double-digit growth rate and the confidence in that growth rate over time, about 2/3 of that business is under contract. Those are, on average, 2.9-year contract businesses -- contracts. There's a high renewal rate, over 90%, for those contracts. So those are all the elements that give us confidence that, that business will continue to grow at that rate.

Perfect. Maybe moving to gross margins. I think the market has been happily surprised, record 49.3% in the most recent guide. But I think as folks take kind of step back and think about where margins are pushing downstream at your customers, particularly memory, wondering why can't the equipment guys get more. And so, I know it's a very different market. You have to add value in order to derive that increase in pricing and margins. But curious, as you sit here, you've talked about trying to raise pricing. How should investors think about that progression? And what are the key underlying drivers that support those higher margins?

Yes, I think that's a great question. First of all, the portfolio becomes more valuable every year. And by the way, investors can look. We report -- now, with the new guidelines on segment reporting, if you look at our segments, for our Equipment business and our Services business, we report our gross margins. You can see those separately. It was 54.5% in recent quarter in our Equipment business.

So the portfolio, as we do our R&D and make our investments in R&D, we think the portfolio becomes more and more valuable every single year. You're working on higher complexity problems for the customers, in some cases providing integrated solutions that basically deliver modular level solutions to the customers, these become more and more valuable. And you should expect the gross margin to improve over time with that higher value.

At the same time, over the last couple of years, we've invested in our pricing process, which means we've got a much more sort of disciplined process in the company to determine the value of the equipment that we're selling to set a target price for the sales team to have the sales team try to achieve that targeted price. That's a much more disciplined process than what we've had before. Our pricing has gone up over the last 2 years with that general improvement in the overall portfolio, and we expect that to continue, C.J.

Perfect. And how does EPIC fit within kind of the gross margin story? Obviously, you're partnering even closer with your customers. You announced this morning a partnership with Micron. I would assume that, that speaks to years down the road of higher margins. But curious today, what's the impact of that investment? And then, how should we think about the progression of margins related to that partnership?

I think investors and customers expect Applied to be working 3 and 4 generations out with our customers. In order to do that, you need lab space. And so what we've built in Sunnyvale, right near our Santa Clara campus, is a large lab to host all of our customers and to do this long-range work with our customers on future technologies. That is really the core of the industry. You have to be 5 to 10 years out in the materials identification and the interaction between those materials in the tools that apply and remove those materials. It takes that sort of long-range investment to work with customers to provide those solutions.

And when we're effective, you get designed in early on. And so you know you have those designs, and in many cases, since they are modular level, those designs can last for generations and generations in process technology. So we needed more space to do that. And C.J. referenced an announcement this morning, one of our impressive customers, Micron, long-standing partner, has announced that they're going to be joining the EPIC Lab, and they'll be part of the collaboration process in the EPIC Lab. So that's a very important announcement for the company. It joined Samsung as another founding partner in the investment.

So we expect to have a large number of our customers in EPIC. It's a very important investment for the company in terms of the future of R&D for the company, and we're looking forward to opening it this fall and fitting it out with the tools that will be the newest generation tools and working with our customers on those.

And so as a plug, SEMICON West, October, I believe you'll be hosting a couple of events around this.

Absolutely. We're hosting an event around the opening of the facility, and it will be exciting and impressive.

Perfect. Perfect. Maybe to close, I think we've got little less than 5 minutes to talk about operating leverage. In late 2025, you had headcount reduction. And you guided OpEx relatively flattish into Q1, which is better than kind of normal seasonality. How should we think about, particularly thinking about a fairly steep recovery, in my view? How should we think about operating leverage for the Applied model going forward?

We expect to deliver operating leverage mechanically by raising spending at a slower rate than revenue growth. That is the expectation. But when you say, well, why is spending going up, most of our growth comes from organic investment. So we are building lab space. We are building manufacturing space. We are adding headcount for this ramp. When we sell more tools, we install and service those tools, so we're hiring the headcount and training the headcount to be ready for the ramps at our customers.

And this quarter, we added a number of new R&D programs that will expand our portfolio in the future, where, of course, we have high confidence in the ROI of those programs. And so with the higher confidence outlook, the longer outlook that we gave to the Street from a revenue perspective, we felt we were in a position to fund those programs.

And from an investor perspective, I would -- if you look at Applied, the way I think about it, we tend to reinvest about nearly half of our profits, and the rest, we distribute to shareholders through buybacks and our dividends. And so you will see Applied continuing to make investments in our core portfolio and in adjacent portfolios. And those are high return on investment products from our perspective.

What's the average cycle time from R&D investment dollar in to ramp HBM revenue out?

I think it's several years. So -- of course, it depends on the application because some applications are modifications of something that's existing, so that can go a lot faster. But if you have something that's 3 or 4 nodes out, that could be a 5-plus-year roadmap.

Perfect. You alluded to capital allocation, half of free cash flow. In the last 12 months, you bought back $4 billion of shares. You do sit in a position where you are above net cash breakeven. I tend to model you around net cash breakeven and return everything. Has your view changed at all? Is there a certain level of cash that you need to run the business? Or is that still a fairly decent kind of philosophy behind modeling capital returns?

I think the way I would look at it from a capital return perspective is we expect to return 80% to 100% of our free cash flow to investors over time. And that's not every quarter, C.J. So in the recent quarters, we've had pretty significant capital investments with the EPIC Lab that we're talking about. We had some other cash needs. So that will fluctuate on a quarter-to-quarter basis. But as you referenced, over the last year, our distributions, I think, were about 86% of free cash flow. So no change there.

Perfect. Maybe to close and thinking a bit longer term, particularly around your EPIC Center investment, how will you monitor or measure success with that investment?

I think that's -- if you look at speed of innovation for the company, that's what we're thinking EPIC will provide us, having our customers in our lab with the brand-new latest development technologies using those, that will speed innovation. The lab itself is built to speed up your ability to install hookup tools, change chambers, do new experiments. The lab is built to accelerate that process. I won't go into the construction details since we have 2 seconds left. But it is designed to speed the process up. It is designed to host customers, and we think that will accelerate our innovation and pace of innovation.

Well, perfect. I think we have to close there. Thank you all for coming, and thank you for coming, Brice.

Thanks, everybody.

Great to see you.

All right. I think we'll get started. I am Shane Brett, U.S. semiconductor equipment analyst. With us today is Dr. Prabu Raja, President of the Applied Materials Semicon Products Group. Let's get started. Dr. Raja , can you start by giving the audience just an instruction and sense of what you do in your role at Applied Materials.

Thanks, Shannon. Yes, about me. Prabhu Raja. I joined Applied years back as an engineer, this is my first job, and I led some of the valuable businesses in Applied Materials like a PVD business and edged and integrated products. I spent lots and lots of time with the customers. Sometimes I think I spend more time in the flights than at home. And the recently I have been to multiple customers through Asia and also the governments right now, governments want to be talking about semiconductor. So the good news is I have a lot of valuable feedback from the customers on the demand, where they are, the challenges. And the bad news is I sleep about 3 hours these days. So if I mess up, just assume it's a jet lag, not me, yes. .

But the Air mode pretty good. Yes. right.

I guess just to get started then as President of the semi products group, just what are your key priorities, say, just where is your focus is with Applied?

So if you really look at those right now, it's all about demand. As I said, I talked to a lot of customers that demand is very high. They all want to know are they getting the tools on time. And I can tell you, pretty much I get e-mails every single day. I want to make sure that we're delivering on time. So we have incredible visibility from the customers. I'm focusing on making sure that we can scale and if we want to deliver the tools on time. That is right, right now. When you really look at the long term, so what do you think about when I'm the slide, what do you think about? I think about this semi-cap market, how big and how sustainable it is, number one. Number two, within semi cap, which part of it semicap is growing faster where we should focus on. Number three, the customers are having the tough challenges right now with all the inflections and how can we work together with the customers in solving those challenges.

These are the 3 high level is a wafer overall semi-cap market. As you can see that the demand is high, the fab utilization is pretty high. We are tracking around 100 fabs today. They're all in various stages, that part is really good. From what I hear from customers, they want -- need to be ready for the super cycle. That's what they call super cycle. I don't know what is the mean. Super cycle. That's why they say that, right? I want to make sure I'm not the bottom line, okay? Number two, if you really look at those, if you segment within those semi cap, not long ago, 1 year back, we say that a, it's a $1 trillion market by 2030. And 1/3 is leading-edge fondry logic, 130 Hiab, 130 memory. That's not changed. Now we are -- potentially, we can do $1 billion -- $1 trillion semi can do $1 trillion this year. Potentially, we can do that. And also, if you really look at the mix, what I see that is the fastest-growing markets are the ones which are enabling artificial intelligence right now, the data centers.

So we really look at those leading edge fondilogic, DRAM, advanced packaging. These 3 are growing faster. And the other ones are slow growth areas, we call it NAND and ICAPs, which are more in general computing. The good news is the fast-growing areas, 3 areas we talked about Applied is #1 in all those 3 areas. And we thought about the areas we invested in those areas. The last 1 is we talk about tough challenges. What are the customers concerned about. If you really look at those, a lot -- the multiple inflections are happening in this area. Each of this year, multiple inflections are happening in each of the areas. So what is customer doing? They are trying to integrate the devices to address their tough challenges. That means you put advanced logic and DRAM together through advanced packaging to solve their challenges. So what we are doing is, we are making it possible with our unique portfolio integrated solutions.

Got it. Got it. And I guess just on this operating environment, so you've talked about just semi cap model, your focuses and then the customer challenges. So we'll focus on the semi-cap model first. You threw out the word super cycle as it relates to sort of AI driving kind of this equipment market. How does that sort of AI drive this equipment market? How is it sort of beneficial to you guys focused on DRAM, logic and advanced packaging?

I think I talked about, I mean, 1 trillion -- 1/3, 1/3, 1/3. That's not there anymore right now. Right now, everything has changed right now. And a lot of right now really that what is a data center. That's all you've got to think of what is enabling a data center. When you really look at those advanced logic and -- sorry, leading edge, let me put the leading-edge logic, advanced packaging and DRAM. These are the fastest-growing markets as compared to that and as compared to the other market, and so the multiple inflections are happening in these areas right now. If you really look at the gate all around. It's extremely complex, and it's going through -- everything is going to 2D to 3D, right? Now you see that that a lot of complexity is there right now. And the complexity means that a lot of more steps, a lot of new steps are coming in more steps. And Applied is in a very good place to solve the complexities right now, okay? That's where I see that where the market is going.

And then I think the good part is we had a strategy, we call it inflection-focused innovations. And we kept talking about it for 4, 5 years, we talk about inflections are coming, coming, coming it's a time. Now we see that. Those wherever we strategize, wherever we invested, that is happening right now. So we are in a very good place in all those 3 areas. So that's why if you really segment it that's the area we are biggest focus right now is paying off.

Got it. So there's inflection focused innovations have sort of led you to DRAM leading logic and advanced packaging and sort of the stars have aligned with AI?

Yes. You are right.

Got it. Got it. I want to talk about 2026 because -- you sort of talked about more than 20% shipment growth for the calendar year. I think if the year plays out as you've sort of called it out to be, that will be the highest growth for Applied since 2021, which was back then the semiconductor industry, just equipment was a big bottleneck at that time with chipset just affecting a lot of other industries. Just how are -- how is the industry an applied prepared relative to 2021 to sort of meet this kind of massive surge of demand?

Okay. The short answer is we are in a much, much better place. I can tell you that. And it's a great opportunity. We are not going to let it go, okay? This is the number one. So you really look at those -- and like I said, I talk to customers a lot. They are giving us great visibility. They are forecasting the demand much, much earlier than ever before. The reason is they want to make sure they're getting the tools. I can see that demand is side, they want to make sure they're getting their tools. More important, when the tools are coming in, these tool needs to be installed. It has to be serviced and for that, you need a manpower much, much earlier, Applied does that.

So you can't just hire a college graduate and ask them to go and do that. These are complex areas you need to learn. So these guys need to be trained for 1 year or so. So for that reason, the customers are giving a visibility much, much earlier right now. That's a great visibility we get. And so that really helps us right now to plan and what do you do with that? -- You're right, COVID was stress test for us. We learned our lessons. And we fundamentally change how we do that in terms of how we plan, how we forecast and how we scale, how we execute, we fundamentally change that. We are in a much better place if you really look at those manufacturing. We can -- the capacity to double compared to what we had before COVID. And since we got a very good visibility from the customers, sometimes 2 years early, even beyond that, and what we do with the visibility is we are going and talking to our suppliers, partnering with the suppliers to early enough. We see where the bottlenecks potentially could be. We are well prepared for it. The suppliers are more ready right now. And the third part is that we talk about service and support. And we are expanding our service team. We are training them.

I think we are in a much better place. Like I said, we are not letting it go this time. For sure, we are more ready right now, yes.

Rallying 2,000 suppliers cannot be easy. So just how have you gone ready for this inflection? I know we kind of spoke about earlier, you guys kind of talk about a 5-year strategy. What were kind of the steps you put in place during the kind of 5 years leading up to this that allowed you to kind of meet this demand?

I think if you really look at those, Applied, if we look on Applied revenue, even after Applied revenue never went down. We didn't come down. We kept in the same every year, we kept increasing that increased the revenue in the last 20 years. That kind of helped us to customers to have more of a confident on us. Other ones we changed the model is right now. I really look at those with the customers, we truly work with our customers as partners rather than just a supplier customer related work with partners. I think we do the same thing for our suppliers. We work with them much more closely. So they're a lot more open, not only their capacity, they are also more open with their subsupplier capacities. So it's all about partnership, how closely work with them, how much visibility you give them this all together.

Got it. It's a tighter partnerships with your suppliers, but also with your customers, I mean you're talking about 2 years plus of visibility, just -- we're seeing 2 years plus visibility, but there's so many tech inflections right now, just this change in kind of rapid pace of complexity. How is that changing the way the equipment industry works with kind of your customers?

Yes. upon industry also talk about particularly Applied materials. Not long ago, I've been in industry for a long time. And they're scaling, if you really look at technologies, not that complex looking back and more predictable or litho-enabled scaling very predictable we had a handful of materials. And so customers had their own process design team. They're pretty much used to tell us, hey, what is my hardware spec, what is my process step. All I have to do is I got to meet the spec. Customers have their own -- the team, which can integrate everything together all steps together. Right now, the complexity has changed. It is really complex right now.

It's more of a customer supplier relationship with all the complexity we are talking about multiple inflections are happening, these inflections are incredibly complex, I can tell you that, like gate all around, you think about that. More than 2,000 steps. A lot of new materials are coming in. And more important, every angstrom matters. I can tell you the good example is in the transistor gate, the small 10-nanometer space. You've got to put 5 to 6 different materials. Everything is 1-nanometer thickness. So that's a 1 angstrom really matter, those kind of complexity. On the top of it, there is the step-to-step when it's out 2,000 steps, some of the steps that are interdependent Okay. So it's more like the Rubik cube that if you touch 1 step, 1 material, it impacts everything. It falls apart. So that complexity is increasing, right? That is more that part of it. On the top of it, the speed, speed is the game. And the customer want time to market becomes important. The customer -- they don't have a time to put together every pieces together.

So fundamentally, customers are working very closely with somebody applied materials. The reason is really look at Applied Materials. We understand the steps better than anyone because we have broad portfolios and with a broad portfolio, we understand the steps and how they're dependent on each other and also we can come with optimized solutes. -- better than anyone. And the customers are coming to us, they're working with us, partnering with us, 3 notes I had, in some cases, 3 to 4 notes I had. So when you have 3 to 4 loads I had and partnering with our customers working together to solve the problem rather than just meeting those hardware spec or process spec, now I become the part of solving the device challenges as a solution partner.

So that gives me incredible visibility on customer challenges, customer road map. I can use that to fund my R&D and more importantly, when you work that early, you are designed in into this. I'm telling you these complexities like a open heart surgery. Once you're designed in, you don't want to touch it, they fall apart. Okay. I think the biggest change is that the relationship what used to be like a supplier customers versus partners.

I think there's a really good segue because you spent kind of billions of dollars to invest in Epic. And kind of my question here is just how should we see kind of the ROI on that billions of dollars that was spent on Topic. You've had some nice customer announcements there. I'd love to hear about that. And just -- should we expect to hear about more customers when are you opening just kind of the road macaron when EPIC is going to come up and up running?

Yes. No, I talked about complexities, right? And also I talked about speed Okay. Right now, really look at those today, if you really look at those, how the innovation walks, innovation, the commercialization, let's call it, how it walks No, from university academic research, if you come with the new materials. Academic research to the equipments to semi manufacturers to the commercial chip takes 10 to 15 years. Go back and look at those finfet and Hikmetalgate, this all inventory somewhere around 10 years, 15 years, but that's the time it 10 to 15 years, it takes.

Right now, nobody has the time to wait for 10 to 15 years with the complexity. Speed is the game. Everyone wants to be the first. So what do they do? Rather than sequential way of innovating and commercialize, what Applied what we are doing is, can we do it in parallel. Can we co-innovate together. That means universities, equipment and customers can we work together in parallel in 1 place under 1 roof, Imagine you come with the new materials you can validate on the equipment, validate on the customer wafers. And your success rate goes up and the time of time to innovation to come has come down dramatically. That's the whole idea of success rate because I can tell you that otherwise success rate is too low because you can come with a new material, it doesn't work in the equipment. Right now we all work together, success rate goes up and also time to market goes down. I think that is the biggest benefit the customers see.

That's the reason that customers want to partner with us, not just customers but on universities and also the peers. They all want to be a part of it. Like you said, we announced with the Samsung. We are talking to and working with other customers, investees, major U.S. university days, also with the peers. And you'll see more announcements in coming months.

Okay. Well, let me word to that then. I guess let's talk about -- dive into the road map here. So leading-edge foundry logic. You sort of highlighted that from FinFET to GAA with backside power that sort of increases your revenue opportunity by about 30% for client fab capacity. Just where specifically, could you kind of give Cardon, where specifically applied is winning for those kind of inflections?

So let me see, when you say leading edge, fondlogic, the we are #1 that process equipment provider in the space right now. If you really look at those, there are 2 big pieces if you really look at those. The 1 big piece is that transistor, other big pieces of wiring. Let me start with the transistors, Okay.

Transistors are the ones I'm talking about the complexity. FinFET, we are #1, if you really look at non-lit, we're very close to 50%, maybe a little less than 50% share right now. That gear Talon makes it even more complex. Like I said, a number of steps are increasing significantly. A lot of new materials are coming in, and these materials are interdependent. They really interact with each other. So the complexity is happening where Applied has a leadership positions, like for example, EP is very critical. And PV is critical, metal yield is very critical. CVD is critical, implant treatment, you can see that where conductor etch -- those are the -- these inflections are happening, complexity is happening, their Applied material has a later sipes.

On the top of it, I talked about interdependencies between the steps. That means if you touch 1 step, it will impact the other step. So that means these all steps need to work together to get the solutions. That means Applied knows how to optimize better than anyone because we know all those pieces, and we know how to connect those pieces to make it work.

An integrated material solutions.

That's what we call it integrated material solutions right now, how we integrate -- in some cases, you can't take it out of the 2 machines. It's all under vacuum. And imagine I talked about 1 nanometer, the 10 atom. -- once imagine you are exposed to yes, it's not -- it's not the same material anymore, right? They all need to be integrated in 1 system. That's what Applied comes in and really helps. And the customer -- what they do is that these are the places we work together 3 to 4 NAND technologies ahead right now because this is not the 1 that you can just come and change it something. So we were 3 to 4 technologies I think it benefits both of us then for customers, they can do early validation on the wafer what we do and also time to market. For us, we get this visibility into the road map, visible into the technologies, road map, challenges, also our R&D funding. I don't have to put R&D in 10 different ways right now. I know I can validate I can fund it in an R&D funding guide of my R&D funding. And even more important, I'm designing. I know 3 notes earlier I'm designed in most cases. So win-win for both of us.

I want to talk about 2 more things within leading edge logic, wiring and just copper. I think the first question is just wiring module and leading-edge chips. How is it changing and where is the opportunity for Applied. And I guess the second part of that question is there's been concerns about PVD copper being at risk from ALD moly, Just what is your sort of answer to that concern?

Okay. Maybe particularly this topic is more personal to me because I joined the company in the PVD wiring. That is my first job, okay. And I've been hearing this. When I joined the company, people say that, hey, copper is going to end, PVDs going to end, don't join that group. That was 25, 30 years back. And yet copper continues to extend and PVD continue to extend. I think -- if you really look at the wiring, is a very valuable and large business for the company right now. And I really look at those in the -- today's GPU, there are 800 plus miles of wiring this is complex. Is look at those. And people -- the complexity in wiring is usually there as transistor. -- people don't give importance to the letting. Wiring is extremely complex. If you talk to the customers, any customer, they'll tell you, Litho is not my bottle like wiring is my bottleneck. That's the first thing they say that, wiring focus on wiring. That is the most important message. So what we see that is now 20-plus layers right now in the GPU right now. So we see that copper -- number of couples are increasing. -- layers are increasing. What is such change is what has changed is to put 1 layer. We used to have 3 different technologies integrated together 10 years back. Today, can imagine, the 7 different technologies are integrated into the 1 system, machine under vacuum to get 1 layer of wiring. Number of layers are increasing, number of technology steps inside each layer is increasing on the wiring. Wiring is big.

And I can tell you that with the breakthrough innovations, at least be our 3-plus nodes of copper rod extend. I can confidently say copper is going to extend for 3 more nodes or even more, okay? We have a path. And that brings to the molycon. There is always a misconception that moly is going to replace copper. Clearly, answer is no. Moly is not going to replace copper. Okay. Copper is going to stay here. Copper will continue to be here, continue to grow and more and more layers. Then moly comes in, really look at those wiring than you hear transistor. Something in between which connects them is contact. There are 3 layers of contacts today -- not -- yes, 3 layers softly contact. Tungsten is the materials -- contact material today Tungsten. That Molly has benefits over tungsten. That's where I see that people are talking about. They -- sometimes we call the contact autocase wiring. That also moly could be replacing tungsten. I think we recently announced a product, the spectral ALD mode. And with the foundry logic, -- we already have a PTR position right now because the ELD mall is replacing tungsten -- it's a very critical rarer it is. And really look at those moly is not as big as what you think really look at the market overall.

Copper is at least a magnitude or more than the mall itself Okay. And copper will stake continue to stay -- and Mali has its 1 place, it could inflect, but copper will be much, much bigger. Our copper will continue to grow. Quite honestly, copper is growing layers are faster than anything. -- the people that doubted you 25, 30 years ago, will be reminded in 5 years, that copper is still around. Copper will be still around. I can tell you that. I told you I worked with the customer 3 to 4 notes. I had, I know exactly what's going on. I do expect.

There you go. Next question I want to move on to DRAM because I think there's a perception that Applied is a bit more of a leading edge, like a foundry logic company, but you guys have gained a lot of share in DRAM over the last decade. Just can you kind of remind the audience where those DRAM share gains have come? Just how you sort of position yourself with this sort of inflection in DRAM spending?

Yes. So 1 thing always, I thought about underappreciated applied is that they think we are a pond-logic company. You're right, not a memory company, really look at DRAM I think over the last 10 years, we increased the market share in DRAM significantly. Where it came from is that patterning conventional patterning, UV patterning in both places, we increased the market share significantly, particularly UV patterning, we increased a lot more than conventional pattern. One thing. Number 2 is that the capacitor module -- that's a pretty complex structure.

We had a lot of key wins because a lot of new materials as a hard mark. We talk about the integrated approach. There is another term we use co-optimized approach. When we said integrated, 1 machine, multiple technologies are all inside the 1 we've seen, not exposed to yes, right? Co-optimize is the 1 that a, they don't have to build the same system, but they interact with each other so that we've got to optimize for each other steps. The step could be a neighboring step or even after 10 steps. Simple example is if you come with a new material, used to know how to deposit the material, if you know how to etch the materials, you should know how to remove the materials is to know how to analyze the material. If you give to the customer on new materials, if they talk to 4 different suppliers to take 7, 8 years, 10 years. They want you to come out with a complete solution for that, hey, give me everything so that I can just insert in my flow. That's where there'll be a lot of caves in capacitor. That is what happened.

The moving forward, let me break down to 2 conventional DRAM and HBMDRAM. Conventional DRAM is following logic road map, very simple. Transistor is going from planar transistor to high-k metal gate to FinFET. So where you get all the learnings? Nobody wants to reinvent the whole thing. Applied has a great share in arch of them, very easy plug and play there, right? Customer work with us. And DRAM also adopting copper layers, number of couple as are increasing in DRAM. So our market share position over many years in foundry logic is really helping us right now. When you go to it and on the top of it, you talk about even convene DRAM, you talk about 4 of score, you talk about 3D DRAM. These are less UV-seUVare more material sensitive. These are the ones, all materials engineering, everything going the third dimension. And this inflection will further help us because it's all a lot more materials, a lot more process technologies, a lot more integrated tools. That's for conventional data.

HPM is very simple, right? HBM, you need 3 to 4x more number of wafers per time -- that's it very simple. So that's where HBM is growing. On the top of it, the HBM packaging part alone is more process intensive. So that's where HBM is growing. It's a combination of inflecting what is happening, combination of number of copper layers growing and the combination of where the HBM is going right now. see. And I think we are really in a good place in all those DRAM influxes right now.

So just to summarize your take on DRAM kind of fam around just materials engineers and mterials engineering is expanding, which is sort of leading to the bigger opportunities for product?

That's correct. Yes.

Got it. I guess on the topic of HBM, at the very start, when you kind of talked about this AI inflection, we kind of called out DRAM, leading-edge logic and advanced packaging. And advanced packaging is is a small portion of your business, but is also a portion of your business where you're very dominant in HBM. Can you talk about just what are applied strength in advanced packaging? And where do you kind of see the opportunities to grow there?

Yes. Like I said, we are #1 in advanced packaging overall. HBr none really 26, what are the strong growth area. HBM packaging and 3D chiplet stacking. There are different names, different customer, different name people -- within advanced package within the even chiplet stacking. People called 3D stacking, whatever it is. So those areas are growing very strong. And Applied has a really good push on there. Really look at this advanced packaging, right? Packaging is nothing but wiring. On-chip firing whatever the learning what we have, we talk about copper wiring, what we had, all the wiring learning, what we had for so many years, all the machines we developed for so many years on wiring, Now all are being adopted off the chip right now. It's nothing but wiring right? So it's a good place to start with for us because all these products are already helping us or the learnings are there. That helps us.

On top of it, we already saw those inflection coming 10 to 12 years back. And personally, I was involved in this. I started this. In Singapore, we started a packaging lab. It's a complete flow lab in Singapore, we call it advanced packaging lab. Sometimes you call it Epic packaging. It does similar to what Epic what we have dream right now. What we do there is we have complete flow. We have customers coming in partnering with us, and to be our peers coming and partnering with us and some of the areas where we are not participating, we are working with them right now. And that gives us a great advantage if customer wants to validate a new skin, they come and work with us right there. That kind of help us.

So we saw that early, we started -- formed a packaging team and this integration flow really helps us right now. The 2 big inflection what is happening right now is bonding for sure. The other 1 is panel, really look at those. The bonding areas, we can see we have partnership with BSE and the panel area, we thought about it even 10 years back 7, 8 years back, there's a Tango system panel, we occurred the company right now. On the top of it, we are developing multiple products in tunnel right now. Since we have a display business for us, it's much easier. Now they have bigger tools to do the planet right now. On metrology, other area of focus on.

I guess my last question would be the sort of I look at this industry and think with market conditions being so tight, you have a once-in-a-generation opportunity to raise prices and improve gross margin.

I know this is coming yes.

What's your view on that?

Now if you really look at those in terms of margin, I think -- ever since Gary joined the company, we increased 7 percentage points margin right now. I think we are on the right track right now. I think we think a little different. I had many questions from you guys, hey, 1 time, can you increase the pricing and all? We think a little differently in this space right now. Right now, I talk about -- we believe in creating the value for the customer and sharing the value for the customer. That's what we believe in. Really look at those, I talk about complexity. I talk about working with our customers much earlier 3-plus notes I had. This all customer, what they care about is device performance, yield, reliability, time to market.

I think what we do is we partner with them, work together with them to address those challenges right now, 3 notes ahead. this is creating a lot of value to the customers. They've got a more chip, they've got a better chips. They are open to share the value right now. We see that already. We see that open to it. You see that complexity is happening where we have high share areas and partnering with the customer much earlier, and we have a control on the value what we can -- what customers can share with us. I think that's the way we think about setting the value. So what do you do with the value, something for a margin and also R&D investment. We want to increase R&D investment, sometimes we want to return to the shareholders in some areas. That is on overall value setting, but let's not forget that. We have a big, big focus on cost.

We talk about costs through technology, cost through innovation, how we can cut down the overall cost. We are working on both fronts. I'm very confident, not only we increase the margin over the years, I think it will continue to incrementally will increase this margin over the years.

Great. Well, on that positive note, that brings us to time. Thank you, Dr. Raja, and.

Thank you.

Welcome to the Applied Materials First Quarter of Fiscal 2026 Earnings Call. [Operator Instructions]

I would now like to turn the call over to Mike Sullivan, Corporate Vice President of Investor Relations. Please go ahead.

Good afternoon, everyone, and thank you for joining today's call. With me are Gary Dickerson, our President and CEO; and Brice Hill, our Chief Financial Officer. Before we begin, I'd like to remind you that today's call includes forward-looking statements, which are subject to risks and uncertainties that could cause our actual results to differ. Information concerning these risks and uncertainties is discussed in our most recent Form 10-K and other filings with the SEC. Today's call also includes non-GAAP financial measures. Reconciliations to GAAP measures can be found in today's earnings press release and in our quarterly earnings materials, which are available on our website at ir.appliedmaterials.com.

Before we begin, I have several calendar announcements. On Tuesday morning, February 24, Applied will host new product briefings at the SPIE lithography and Patterning Conference in San Jose, and you're welcome to join us in person. We are also very pleased to announce the first event in our 2026 master class series. We plan to cover transistors and wiring on Wednesday, April 8, at 9:00 in the morning Pacific Time and will host a live webcast.

Finally, we're excited to announce 2 special events during [indiscernible]. On Monday afternoon, October 12, we plan to host an investor open house event at the new Epic Center in Silicon Valley, California. We hope you'll join us there. And on Tuesday morning, October 13, we plan to host an investor breakfast in San Francisco. We hope to see you in person or on the webcast. And with that introduction, I'd now like to turn the call over to Gary Dickerson.

Thank you, Mike. In our first fiscal quarter of 2026, Applied Materials delivered revenue and earnings above the midpoint of our guided range. Our strong performance and outlook for 2026 and beyond are fueled by the acceleration of investments in AI computing. AI is at a tipping point where improvements in performance and cost translate to real-world applications that deliver meaningful productivity gains and return on investment for users. The race to build out AI infrastructure is driving unprecedented spending on semiconductors, semiconductor manufacturing capacity and research and development.

Today, the most critical and fastest-growing markets are leading-edge logic, high bandwidth memory DRAM and advanced packaging. These are areas where Applied has strong leadership positions as well as an innovative pipeline of solutions to enable next-generation technologies. In my prepared remarks, I will provide my perspective on how our markets are evolving, explain the role we are playing to enable the AI road map and describe how we're turning the opportunities ahead of us into sustainable, profitable growth for Applied. Semiconductors are the heart of the AI technology stack. With the accelerated growth of AI end markets, we believe that global semiconductor industry revenues can potentially reach $1 trillion in 2026, several years earlier than prior predictions.

For Applied, we expect to grow our semiconductor equipment business more than 20% this calendar year. We see the demand profile weighted towards the second half of the calendar year with availability of customer clean room space being a key factor pacing the rate of investment. Our largest customers are giving us increased longer-term visibility to ensure we have operational capacity and service support in place for their ramps. Based on this visibility, we expect strong growth momentum to be carried into 2027.

Improvements in AI data center performance and cost directly impact AI adoption rates by increasing the number of tokens generated per second and lowering the total cost of ownership, which is dominated by energy consumption, more AI applications become economically viable. This need for higher performance and more energy-efficient AI computing is reshaping semiconductor industry investments and driving high growth rates for Applied in leading-edge logic, HBM DRAM and advanced packaging. In leading-edge logic, our customers are adding capacity at FinFET nodes while simultaneously ramping gate all around nodes. Applied is the clear #1 process equipment provider in leading-edge logic with strong leadership positions across materials deposition, modification and treatment as well as in conductor etch and e-beam technologies.

Gate-all-around nodes grow our available market considerably while also providing a catalyst for multiple points of market share gain. In DRAM, customers are aggressively adding capacity at 6F-squared nodes, while in parallel, developing next-generation DRAM device architectures. AI computing is driving significant demand for high-bandwidth memory DRAM, which has larger die sizes and requires 3 to 4x more wafer starts per dollar bit than standard DRAM.

In addition, the number of dies in the HBM stacks is increasing from 12 today to 16 and then 20 or more in the future. This further grows demand for wafer starts and advanced packaging. Applied is the #1 process equipment provider in memory today, thanks to our very strong position in DRAM, where we are the clear leader in materials deposition for both wiring and patterning as well as conductor etch and e-beam technologies.

In advanced packaging, the mix of customer investments is changing. We expect the fastest-growing market segments in 2026 to be HBM and 3D chiplet stacking. Applied has very strong market share in both of these areas, thanks to our leadership and materials deposition and removal technologies, which is enabling us to extend our overall leadership in advanced packaging. In NAND, we are forecasting modest growth in equipment demand in 2026 and and we expect NAND to remain less than 10% of wafer fab equipment spending.

In ICAP, customers who serve the IoT, communications, automotive, power and sensor markets, we expect wafer fab equipment to be approximately flat year-on-year, both globally and in China. At Applied Materials, our strategy is inflection-focused innovation. By focusing our R&D resources on the development of high-value solutions to enable major device architecture inflections, we are accelerating our customers' road maps and driving sustainable value capture and margin expansion for Applied.

Our innovators have generated an expansive pipeline of next-generation products. products we have released over the past several years are significantly contributing to our growth in 2026. One example of this is our unique cold field emission e-beam technology where we expect revenues to double this calendar year to more than $1 billion and support process diagnostics and control, being one of our fastest-growing businesses in 2026.

In addition, our leadership in CFE e-beam imaging accelerates learning rates for next-generation chip architectures and adoption of Applied's process equipment portfolio. In 2026, we are planning to launch more than a dozen new products, including 3 for advanced logic and DRAM, which we announced earlier this week. Our [indiscernible] radical treatment system delivers item-level precision engineering of nanosheet surfaces, which enables higher speed next-generation gate-all-around transistors. [indiscernible], which is the newest variant of our [indiscernible] etch platform enables angstrom level precision for critical edge steps in gate-all-around transistors and advanced DRAM.

This extends our conductor etch market leadership in advanced logic and DRAM. And our spectral ALD system enables selective deposition of monocrystal [indiscernible]a new material that can reduce contact resistance and advanced logic devices by up to 15%. We Applied is leading the transition from ALD Tungsten to ALD Moly and logic contact formation. As transistor counts and wiring layers increase, copper PVD steps will continue to grow significantly, remaining an order of magnitude larger than ALD moly.

This week, we also announced our first EPIC co-development agreement, Samsung Electronics. Applied's global Epic platform is designed to support high velocity co-innovation with our customers and R&D partners. For chip makers, EPIC will provide significantly earlier access to Applied's R&D portfolio, enabling faster cycles of learning and accelerated transfer of next-generation technologies into high-volume manufacturing.

For Applied, EPIC will give us better multi-node visibility to guide our R&D investments while improving R&D productivity value sharing and our ability to be designed in with our products and advanced service technologies. As our customers ramp complex new devices in high-volume manufacturing, we see accelerating demand for our advanced services, supporting a double-digit growth rate for our service business. We have many valuable service innovations that accelerate the transition of new technologies from lab to fab and then accelerate ramps to enable increased yield and output in high-volume production.

Several years ago, we launched Applied's proprietary AIX or actionable insight Accelerator software capabilities. Today, we have more than 30,000 chambers connected to AIX servers using AI-powered monitoring, diagnostics and analytics. Across these connected tools, we are seeing 30% faster response times, enabling increased wafer output for customers and better service engineer productivity for Applied. In addition, we now have automated all our major distribution centers with state-of-the-art AI-enabled robotic systems, significantly improving our parts delivery speed and accuracy as well as inventory optimization.

Before I hand over to Brice, let me briefly summarize. Semiconductors are the heart of the AI technology stack and as AI adoption accelerates, we see industry revenues potentially reaching $1 trillion in 2026. Our inflection-focused innovation strategy is generating an expansive pipeline of higher-value products that are extending Applied's leadership in leading-edge logic, memory and advanced packaging and enabling us to grow our semiconductor business more than 20% in 2026. And we are driving deeper co-innovation engagements with our customers to enable energy-efficient AI architecture inflections and accelerate the transfer of new technologies into high-volume manufacturing. Epic will come online later this year bringing together key innovators from our customers and applied to significantly increase value creation velocity. Brice, over to you.

Thank you, Gary, and thanks, everyone, for joining us today. I'm pleased that Applied delivered strong first quarter results, and I'd like to thank our global teams and partners for their contributions. Looking ahead to Q2, we anticipate strong growth in our semiconductor systems business along with healthy gross margin and increasing profitability for the company. As Gary described, our business is strengthening with positive demand indications throughout the ecosystem. We are tracking higher levels of planned CapEx from cloud service providers. Semiconductor factory utilization is rising across all device types. Leading-edge foundry logic and DRAM capacity is essentially full and prices have increased. These dynamics are driving significantly longer visibility from our own customers who are increasing the number of new factory projects and fab expansion scheduled to be completed over the next several years.

As the largest supplier of process equipment and leading-edge foundry logic DRAM and advanced packaging, A major priority for us is ensuring we have the capacity to support our customers over this period. Over the past several years, we've nearly doubled our system manufacturing capacity and strengthened our supply chain operations. In recent quarters, we've given our direct suppliers longer visibility into our requirements, which has allowed them to proliferate the demand signal throughout the supply chain to secure the materials and labor needed to support growth.

At Applied, we've proactively increased our inventory by nearly $500 million year-over-year to meet the increasing build plans. As a result, we are well positioned to meet the increasing demand we are seeing from our customers. Next, I'll briefly summarize our Q1 results. Revenue of $7 billion was in the upper end of our guidance range and down 2% year-over-year. Revenue in China declined 7% year-over-year and represented 27% of combined semi equipment and AGS sales and 30% of overall sales. Non-GAAP gross margin was 70 basis points above the midpoint of our expectation and grew 20 basis points year-over-year to 49.1%.

Non-GAAP OpEx was in line with our expectation and grew 2% year-over-year to $1.34 billion, with an 8% increase in R&D investments largely offset by G&A spending reductions. Non-GAAP operating profit declined 4% year-over-year to $2.1 billion. Non-GAAP earnings per share of $2.38 was at the top of the guidance range and flat year-over-year. Included in our GAAP results is an accrual of $252.5 million related to an export controls compliance matter we disclosed in our 2022 10-K and later filings. The Department of Justice and SEC have closed their inquiries into this matter with no enforcement actions. We issued a press release and filed an 8-K related to the seat we reached with the U.S. Department of Commerce Bureau of [indiscernible], to resolve its inquiry and you can find all of our comments related to the matter in these documents.

Next, I'll summarize our segment results. As a reminder, beginning in Q1, our 200-millimeter systems business has moved from the Applied Global Services segment to the Semiconductor Systems segment. Corporate support costs previously reflected in Corporate and Other are now being fully allocated to our businesses. and display business results are now included in other. To help with your models, today's earnings press release and slide presentation include tables that provide a recast for these changes. Semiconductor Systems revenue exceeded our expectation in Q1 and included record DRAM revenue. On a year-over-year basis, revenue declined 8% to $5.14 billion.

Non-GAAP gross margin increased 100 basis points to more than 54% driven by our continued focus on value-based pricing, along with manufacturing cost improvements. Non-GAAP operating margin declined 80 basis points to 32.9%. Applied Global Services delivered record revenue of $1.56 billion, which exceeded our expectations and grew 15% year-over-year. Non-GAAP gross margin increased 210 basis points, and non-GAAP operating margin grew 320 basis points.

Turning to the balance sheet. We generated $1.69 billion in cash from operations. Free cash flow of $1 billion included elevated capital investments as we made continued progress building the Epic R&D center in Silicon Valley and further expanded our systems manufacturing capacity. Finally, we distributed $702 million to shareholders in cash dividends and stock buybacks. Over the past year, we've distributed over 85% of free cash flow to shareholders.

Now I'll share our guidance for Q2. We expect company revenue of $7.65 billion, plus or minus $500 million, which should be up around 9% sequentially. We expect non-GAAP EPS of $2.64 and plus or minus $0.20. Within this outlook, we expect Semiconductor Systems revenue of around $5.8 billion, AGS revenue of about $1.6 billion and other revenue of around $250 million. We expect non-GAAP gross margin to increase to approximately 49.3%. We expect non-GAAP operating expenses of around $1.415 billion. Given increased visibility and confidence in the industry's growth outlook, we are accelerating R&D co-development projects with our customers and partners. We remain highly focused on growth, productivity and margins. Finally, we are now modeling a non-GAAP tax rate of around 11%.

In summary, the investments we've made over the past several years have put us in a terrific position for profitable growth. The global expansion of AI infrastructure is translating to accelerating demand for our most enabling products in leading-edge foundry logic DRAM and advanced packaging. Along with advanced services that help our customers accelerate ramps and yields. Working closely with our customers, we're increasing the energy-efficient performance of logic chips, compute memory and systems and we are sharing in the value we create.

Applied is the materials engineering leader in the fastest-growing segments of the semiconductor market, and we are driving our operations teams and supply chain partners to increase capacity and output as our customers ramp in 2026 and 2027. Thank you for listening. And now, Mike, let's begin the Q&A.